Eyebrow Shapes for an Oblong or Long Face: Tail, Arch and Brow Height

Which face is this page for — oblong, long, or long and thin?

All of them, because they are one question wearing several phrasings. People who have this problem type it as an eyebrow shape for an oblong face, an eyebrow shape for a long face, and — least often but most specifically — an eyebrow shape for a long thin face. The advice given back does not distinguish between the three, and neither does this page.

This site measures the outline as face length divided by cheekbone width, hairline midpoint to chin over cheekbone apex to apex. On that measurement the published rule set here centres an oblong face at 1.65, with a band from 1.47 to 1.83, and centres an oval face at 1.45, with a band from 1.27 to 1.63. If you have never taken the two numbers, the calculator on this site will do it, and section eight below explains the one place where the distinction actually changes the brow answer.

The phrasings do not mean the same measurement, and the record shows this directly. In 2023 someone described higher brows as seeming to make the middle third of the face look too long, and the answer they received had to stop and ask what the middle third was meant to cover — the nose area? Total length and the span from brow to nose base are different distances. The brow sits on the boundary between them, which is exactly why it can change one without changing the other, and why an instruction aimed at total length can land on the mid-face instead.

The thin part of the third phrasing is worth a warning. Two of the seven people in the record who called their own face long also called it full or chubby in the same sentence. Neither had a long thin face, and both were given brow advice anyway. Length relative to width is a ratio; whether the lower face carries weight is a separate fact, and no source here separates them.

What does the record agree on?

Density, and only density. Five of the seven contain the same instruction in different words: put the hairs where they already are, upright and outward, and hold them there. Set them with a clear gel. Set them with soap and a spoolie, which holds the hairs upright without adding anything. Brush up and out so they find their natural direction. Comb them into shape. One source recommends a growth serum instead, which is the same goal reached by a different and less reversible route.

The reason this one setting has no dissent is structural, and it is worth stating because it explains why the other settings do have dissent. Brushing and setting adds no hair and removes none. Every other instruction in the record either takes hair away or puts pigment in. An instruction that changes nothing permanently has nothing for anyone to argue with.

That does not make it trivial. Two of the sources that recommend it are addressing brows that are sparse rather than misshapen, and sparse is a density problem that no shape can solve — hair that is not there cannot be carved into an outline, only drawn toward one. Drawing individual hairs at the inner start is also the only move in the record that changes the brow's proportions without touching its outline, which is why it comes back in section ten as one of two things the record independently settled.

The serum is the exception to the reversibility, and it arrives with the only pharmacological caution anywhere in the record: avoid growth serums containing prostaglandin, because that class causes fat loss around the eyes. The caution names something real. Periorbital fat atrophy is a documented sign of prostaglandin-associated periorbitopathy, it is reported for the whole class of topical prostaglandin analogues rather than for one drug, and it partly to wholly reverses when the drug is stopped.

The documented effect belongs to a class of eye drops. Whether an over-the-counter brow serum carries the same risk at the same site is not established by anyone in the record, and it is not established here. What can be said is that whoever raised the caution did not make the distinction either, and that a caution which is right about the drug class and silent about the dosage form is still worth reading on the label before buying.

One more asymmetry, which no source in the record mentions. The oldest source here offers the same person three different density routes in one exchange: set the hairs with soap, have them drawn in with a fine pencil, and have pigment implanted by a professional. The first is undone by washing your face. The second is undone by the same wash. The third fades on the technician's schedule and not on yours. They are presented as a list of options with nothing said about the difference between them.

Straight brow or arched brow: why the argument never closed

The oldest source in the record contains both sides. One says an oblong face suits straight brows best, or straight with a slight upward slant, because a brow that goes down at the end makes the face appear longer. Another recommends having pigment implanted with a slight arch. Neither refers to the other.

After that the arch side keeps appearing and the straight side nearly disappears. A subtle arch will lift the face more. Arch the brows so that eyes, lips and brows take up more visual space and the face reads as more even. Brush up and out to help the hairs define an arch. Then, three years and three months after the first exchange, one source says not to arch from the top and the bottom so much — which is a reduction instruction, not a straight one, and lands between the two positions rather than on either.

Here is the part that has to be marked as this page's reading rather than anyone's claim. The word arch is doing two jobs in the record. It names how much the brow bends between its inner start and its tail. It also names which direction the tail runs in. Those are independent settings: a brow can be nearly straight and still have a tail that drops, or strongly bent and still finish level.

Sort the six positions by which job they are actually doing and the argument shrinks. Four are about bend. Two are about direction. They were never compared with each other, because no source in the record distinguishes the two — and the one instruction that does state a mechanism is a direction instruction wearing the word arch. A tail that goes down continues the face's vertical line past the eye and extends it. That is a claim about where the tail ends, and it says nothing at all about how much the brow bends.

So the bend argument has no mechanism anywhere in the record. Lift and open up the face appear as promised results, never as reasoning, and no source says what it is about a bend that would shorten a face that is long relative to its width. The direction argument does have one. That asymmetry is the closest thing to a verdict this subject has, and it comes from reading the six positions against each other rather than from any of them.

One caveat on the source of the direction rule, and it is not a small one. Whoever gave it says so in the same breath as saying they could not see the face. The only instruction in the record that ties a brow direction to this face shape by name was written by someone working from the words of the question alone. That does not make it wrong. It does mean that nobody in four years and nine months has supplied a reason for it beyond the one sentence, and that the sentence is doing all the work.

Shorten the tail, or is that a photograph problem?

The shorten side has four entries. The length of the tails makes the eyes look droopy when they are not. Trim the eyebrows a bit. Too long over the nose and a little thick. A bit too long and going into the middle by the nose. Note that the last two are about the inner end and the first two about the outer tail — opposite ends of the same brow, described with the same phrase, by different people in different years. Nobody in the record separates them, and they are not the same instruction.

The lengthen side has one entry, and it is the most recent thing in the record: extend the end of the eyebrows. Between the two sits the longest single argument anywhere in the pool, and it changes the subject in a way that dissolves the dispute rather than settling it.

The argument is that horizontally short brows are not a look. They are a byproduct. Tail hairs are the sparsest part of the brow, and they are the first thing a technician tweezes whenever a shape is changed in the chair, so a brow that gets reshaped several times a month loses its ends whether or not anyone chose that. Photographs of professional brows are photographs of that byproduct. People whose brows are not reshaped constantly can carry a long tail, and the mistake is reading a reference image as a target.

That is this page's reading of what the argument does, not a claim made in it: on this account the two sides are not disagreeing about long faces. They are comparing the person asking to different images. One side is comparing her to a face. The other is comparing her to a chair.

The same source supplies the measuring rule, and supplies it as a defence of a long brow rather than as a reason to cut one — the inner edge lining up with the nostril and the outer edge with the outer corner of the eye, offered as evidence that the length is already right. Elsewhere in brow advice that rule is used to decide where a brow should stop. Used that way it becomes a shortening rule, and the tail gets cut to reach a line it was never short of.

What the cut costs is the part the record states only once. Someone describing themselves as over fifty says to leave the brows alone, because waxing or plucking them long enough can stop the hair returning as it did, or returning at all. It is the only place in the record that supplies a time horizon, and it is the reason this dial is not free in a way that brushing is.

And nobody who said trim a bit gave a length, so here is what the words are worth. Five millimetres off a 55 mm brow is 9.1 percent of it and 0.197 in. Ten millimetres is 18.2 percent and 0.394 in. On a 65 mm brow the same five millimetres is 7.7 percent. The instruction is identical in all three cases and the change is not.

Raise the front: prescribed, reported as backfiring, prescribed again

July 2022: the brows sit a little low toward the nose, so pluck a few hairs from underneath there and raise them a little, which lets the eyes stand out more. February 2026: brush and groom them, and maybe thin them out a bit from the bottom side so they are not as low. Three years, seven months and twenty days apart, to two different people, in nearly the same words and with no number in either.

Between them sits June 2023, and it is the only place in the record where the instruction is examined from the receiving end. The person asking there already has high brows, likes how they look, and writes that they seem to make the middle of the face look too long. One answer says the height is right as it is and that the thing to change is the middle — let the brows grow in a little there. Another stops to ask what the middle third is supposed to mean, the nose area? Nobody argues with the report. Nobody refers to it either.

The second prescription postdates that report by two years, eight months and eighteen days. It does not mention it, and it does not need to, because there is no shared measurement for the two to disagree about. That is the mechanism by which an instruction survives being reported as a failure: the failure was described in adjectives, the instruction was given in adjectives, and there is nothing for either to be checked against.

Here is the size of the lever, worked. Removing two millimetres of hair from under the front of the brow raises the front edge by two millimetres, which is 0.079 in. Against a face 190 mm long that raise is 1.05 percent of the total length. If the front of the brow sits 60 mm below the hairline, raising it three millimetres moves it from 31.6 percent of face length to 30.0 percent — a change of 1.6 percentage points. Both the 190 mm and the 60 mm are stand-ins; substitute your own two numbers before using any of it.

The point of the arithmetic is not that the change is too small to see. Two millimetres at the front of the brow is visible. The point is that the instruction delivered with the most confidence in the record moves the smallest measured quantity on the face, and it is the only one of the three dials that is delivered by removing hair that the record itself says may not grow back.

Making level and a little into numbers you can check

The record's entire vocabulary for the brow is qualitative: a little, a bit, slight, subtle, some, up and out, not as low. Six hundred and thirty separate pieces of advice were read across the seven sources, and three of them contain a number about a brow. One is an angle. One is an interval. One is a duration. Not one is a length, and not one is a height.

The angle is a right angle opened to about 120 degrees, offered as a way of kicking the sides of the brows out. Nobody says which corner is being measured, or from what. Read the only way it can be read without inventing geometry — the tail's direction rotating by 30 degrees — the tip of a 25 mm tail travels 2 × 25 × sin 15°, which is 12.94 mm, or 0.51 in. On a 20 mm tail it travels 10.35 mm, or 0.41 in. Hold that against the drop table below: the single number the record offers for the brow itself moves the tail tip further than the entire distance between a level tail and a visibly down one on the same run.

The interval is five weeks, which is 35 days, given by someone describing their own cleanup routine and what happens without it. It is not offered as advice for a long face and it is not tied to any shape. The duration is roughly thirty seconds, for lightening the hairs — a colour move rather than a shape move, and the only time figure in the record.

So the numbers below are computed here, not collected. One inch is exactly 25.4 millimetres and has been since 1959, which makes every conversion on this page a single division with nothing rounded twice. Five lines are worth measuring, and each one decides a different argument.

Brow length, from the inner start to the tip of the tail, following the brow: a 55 mm brow is 2.17 in, a 60 mm brow is 2.36 in, a 65 mm brow is 2.56 in. Tail run, the horizontal distance from the highest point to the tip — this is the input every drop figure depends on, and nobody who discussed the tail in the record asked for it. Tail drop, how far the tip falls below the level of the inner start, negative if it finishes higher. Inner gap, between the two inner ends. Lid gap, from the upper lash line to the underside of the brow where that gap is narrowest.

Worked, on the tail. Over a 25 mm run, a level tail drops 0.00 mm. Five degrees down drops 25 × tan 5° = 2.19 mm, which is 0.09 in. Twenty degrees drops 9.10 mm, or 0.36 in. Thirty degrees drops 14.43 mm, or 0.57 in. The whole spread between level and clearly down is therefore about nine millimetres on that run — and the 30-degree rotation from the record's one angle moves the tip 12.94 mm, which is more than all of it.

Worked, on the inner ends. Moving each inner end two millimetres inward narrows a 20 mm gap to 16 mm — 0.79 in to 0.63 in, a 20 percent change produced by four millimetres of drawn hair and no removal at all. Three millimetres per side gives 30 percent. On a narrower 16 mm start the same two millimetres per side is a 25 percent change. The starting gap here is a stand-in; measure yours.

Worked, on the front. Raising the brow's front edge by two millimetres turns a 10 mm lid gap into 12 mm — 0.39 in into 0.47 in — and on a 190 mm face it is 1.05 percent of the total length. Three millimetres gives 13 mm, or 0.51 in, and 1.58 percent. Those two lines are the whole size of the instruction that section five traces across three years and seven months.

Drop = tail run × tan(tilt). The run is the horizontal distance from the highest point of the brow to the tip of the tail; 25 mm is a stand-in input, not a target, so substitute your own. Every inch figure is the millimetre figure divided by 25.4, exact since 1959, rounded to two decimals. At a 20 mm run the same four tilts give 0.00, 1.75, 7.28 and 11.55 mm — so the run matters more than the angle, which is why nobody in the record who talked about the tail ever asked how long it was.

The tables above are the short version. The full sheet — the five measurement lines with a diagram, the tail-drop sweep from level to 30 degrees, what a trim of five or ten millimetres is worth as a share of the brow, the inner-gap arithmetic, the lid-gap arithmetic, and a test log sorted by what each step costs to undo — is available as a file you can print or keep on your phone: Download the oblong-face brow chart (printable) ↓

What the advice addressed instead of the brow

The largest source in the record is a question that asked for makeup or eyebrow advice, and what it received most of was contouring and curtain bangs. Neither is a brow. That is not an error, and the second of the two is not a change of subject either: it is what follows from the only mechanism anyone in the record actually states.

The mechanism is this: a fringe cut straight across the forehead makes a hard horizontal line, and strong brows make a second one directly below it. Two parallel horizontals stacked at the top of the face read as a boundary, and everything below the boundary is then the length of the face. Growing the fringe out removes the upper line and draws the eye down toward the middle. On that account, full brows do not balance a long face; they mark where it starts. It is a single answer among hundreds in that source, and it is the only explanation anywhere in the record for why adding density to this face could make the length worse rather than better.

It also runs directly opposite the arch instruction, which promises lift. Both cannot be mechanisms of the same thing, and no source in the record notices that they are in competition. The difference is that one of them says what it is doing to the eye's path and the other only says what the result will be.

The rest of what was offered instead of the brow, in the order it appears: whitening the teeth plus contour along the hairline and temples; light contour at the temples and jawline to compress the face; a video search aimed at whichever single feature bothers you; having the brows shaped professionally; and reducing body fat. The last of those came attached to the only answer in the record that could not decide whether the face was oblong or oval, and it is a different mechanism again — not a proportion argument but a composition one.

Two answers in the January 2023 source take opposite positions on the shape itself. One writes that this face shape is treated as a deficit and that the faces praised are shorter and rounder. Another writes that the same shape reads as intelligent and confident. Neither is a measurement, neither is attached to a brow, and nothing on this page adjudicates between them. The site's own position is the one it states everywhere: there is no better or worse outline, only choices that work with the one you have, and a brow decision made from that argument is a decision made from nothing.

The same source contains the only reported outcome of a structural change in the entire record. Filler was placed to add width to an oblong face; the report is that it did not seem to work well; the next sentence is a plan for jaw surgery. One sentence long, nothing measured, no date given for the follow-up. It is the single most informative thing in the record about whether the outline can be changed rather than styled, and it is worth exactly one sentence.

Two gaps close this section. Two of the seven people asking named hooded lids — one confirming it when asked, one listing it among their features — and neither received a brow answer for them. The first got no further answer on the point at all; the second got eyeliner advice, which is about the lid and not about where the brow sits over it. And the most recent source in the record drew a single answer. Advice on this face shape is not only unmeasured. It is thin.

Does the answer change if your face is oval rather than oblong?

The one answer in the record that could not choose between the two shapes then gave both the same instruction, which is either careless or correct, and the rule set says which. This site's published bands are 1.47 to 1.83 for oblong and 1.27 to 1.63 for oval. They share 1.47 to 1.63. Inside that overlap the shape name is not a fact about your face so much as a report of which side of a shared band you landed on, and this site reports a second shape rather than forcing one for exactly that reason.

Where the distinction does bite, it bites on the arch. This site's oval-face eyebrow page finds the arch instruction reversing across the overlap and puts the reversal near 1.55, with straight brows favoured above it. That reversal is an oval finding, and it needs the oval band underneath it. Above 1.63 there is no oval reading left to reverse, and what remains is the only question in the record that has a stated mechanism attached: which direction the tail runs in.

So the practical answer splits cleanly. Below 1.63, take the ratio before taking any instruction that begins with a shape name, because two opposite sets of advice are both addressed to you and the number decides which. Above it, the bend question drops out and the tail question is what is left.

One thing does not change anywhere on that scale. Every position in the record about density is shape-blind — nobody who recommended brushing the hairs up and out first asked what the outline was, and nobody who recommended setting them needed to. It is the only instruction on this page that you can act on today without measuring anything, and that is a large part of why it is the one to act on first.

Nearly five years of advice on this face, in the order it was given

Seven sources, four years and nine months between the first and the last, taken in the order the advice was given rather than the order it was found. The fourth column is the one to read down: what the rest of the record did with each position afterwards.

Five phrasings a person with this problem would actually type were searched, in twelve searches: oblong face and eyebrows; an eyebrow shape for a long face; the best eyebrows for an oblong face; eyebrows for a long thin face; and eyebrows making the face look longer. Seven sources qualified. To qualify, a source had to be a real person describing this face shape or a real person answering one, had to carry a retrievable date, and had to contain an instruction about the brow rather than only reassurance. Six hundred and thirty separate pieces of advice were read across the seven. Excluded, and the exclusion is the interesting part: the three largest bodies of brow advice turned up by the same searches contain no long-face material at all and are not in this table, so what follows is not the loudest record available, only the on-topic one. Also excluded: material whose only shape talk was surgical; one source on an oval-or-oblong face with full cheeks whose brow advice is already used on this site's oval-face eyebrow page; and anything with no retrievable date. Sources are described here by who was asked and what was said, and are never linked.

What the record settled, and what it kept arguing about

What it settled, twice and independently. Bring the inner ends closer. October 2021 says to draw small individual hairs at the inner start of both brows so the two ends sit closer together in the middle. June 2023, one year seven months and twenty-six days later, independently and to a different person, says to let the brows grow in the middle a little more. Same horizontal edit, arrived at separately, and between them the only two positions in the record that change a brow without taking hair out of it.

What it settled by default. Density, for the structural reason given in section two: it is the one setting that adds nothing and removes nothing, so there is nothing for anyone to argue with. Five of the seven recommend some version of it and none of the seven disputes it.

What it kept arguing about. The tail's length, across four years and nine months, unresolved — and from January 2025 onward the argument is not even about faces any more, because the strongest entry in it is a claim about what a professional reshaping chair does to the sparsest hairs at the end of a brow. The bend, which was never separated from the tail's direction, so it is not clear that there was an argument to have.

What it never examined. The height. Prescribed in 2022, described from the receiving end in 2023, prescribed again in 2026 in nearly the same words. The description was not referred to by the second prescription, and could not have been, because neither side had a number to refer to.

Six things the record does not contain, which is the honest way to say what is still missing from this subject.

A length. Three numbers across the 630 pieces of advice read — an angle, an interval and a duration — and not one of them is a length or a height in any unit. Every instruction about how much to trim, how far to raise and how short is too short is given without one.

A request for the ratio. Not one answer in the record asks how long the face is relative to how wide it is, which is the only question that would separate oblong from oval. The single answer that could not separate them gave both shapes the same instruction, and section eight is the only place in this page where that turns out to be defensible.

A before and after with a measurement attached. The one structural outcome reported is a single sentence about filler that did not seem to work well, with no date for the follow-up and nothing measured at either end.

An answer for hooded lids. Two of the seven people asking named them. One got no further answer on the point; the other got eyeliner advice, which addresses the lid and not the brow sitting over it.

A distinction between a tail that finishes level and a tail that finishes high. Straight, level, slight slant up and arch are used for each other throughout the record, which is why the bend argument and the direction argument cannot be separated from inside it.

A maintenance interval tied to a shape. The only interval in the record is five weeks, offered as one person's habit and what happens without it, not as advice for a long face or any other outline.

One last note on the record's precision, which is the opposite of what it looks like. The 120 degrees is the most confident number anyone offered for the brow itself, and it is unusable as given: no corner is identified, no reference line is named, and the two natural readings put the tail in different places. What survives of it is the 30-degree rotation, and that rotation moves the tip of a 25 mm tail by 12.94 mm — more than the whole distance between level and visibly down on the same run. The record's only number for this face is larger than the disagreement it was offered to settle.

• Set before you shape, and undo it the same evening. Brushing the hairs up and out and holding them with a clear gel is the only step in the record with no dissent behind it, and it washes off. Nothing else in this list does.
• Measure the tail drop before you believe too long. Level, or how far below level, in millimetres, then convert. On a 25 mm run the distance between level and twenty degrees down is 9.10 mm, or 0.36 in. If your drop is inside that, the instruction was never about your face.
• Work on the inner ends before the height. Two millimetres per side narrows a 20 mm inner gap to 16 mm — 0.79 in to 0.63 in, a fifth of the span — and removes nothing. Raising the front by the same two millimetres costs hair that the record says may not return.
• Change the tail's direction first and its length last, if at all. Direction is a setting move; length is a removal, and the sparsest hairs on the brow are the ones at the end, which is exactly where removal starts.
• Treat every recommendation that involves a professional chair as a different category of decision from the rest of this list. Repeated reshaping is how the tail thins on the record's own account, and a shape chosen in a chair is chosen against the chair's reference images rather than against your face.
• Leave pigment and any growth serum until everything above has been tried, and read the serum's actives before buying it. The class the record warns about is documented to cause fat loss around the eye when used as a drop; whether a brow serum carries the same risk is not established anywhere, including here.