Midpoint Calculator – Midpoint, Distance and the Missing Endpoint

Your points, on the grid

Numbers are tidy, but a picture is the proof. The plot redraws itself with every keystroke — watch the diamond stay exactly halfway between the two dots, however far you drag them apart.

The coordinate-geometry neighbours

A midpoint rarely travels alone — these are the tools it is usually paired with on the same homework page.

How it works

One idea — averaging — does all the heavy lifting on this page, in both directions.

The math behind the default example

Two ideas worth slowing down for

Why an average can be a place — and where midpoints quietly run the rest of geometry.

The third coordinate: midpoints in 3D

The midpoint formula does not care how many coordinates a point has, because it averages each coordinate separately. In three dimensions the midpoint of (x1, y1, z1) and (x2, y2, z2) is ((x1 + x2) ÷ 2, (y1 + y2) ÷ 2, (z1 + z2) ÷ 2): the same averaging, with one more coordinate joining in. The Advanced options panel on this page works it out for you — enter the z-coordinates of your two points and the 3D midpoint appears alongside the 3D distance, so the two-dimensional result above and the three-dimensional result stay consistent with each other. With the defaults, both z values are 0, so the 3D midpoint is (4, 6, 0) and the 3D distance is the same 7.21 as the two-dimensional distance: a flat segment lying in the xy-plane has no extra length to add.

The distance formula grows the same way. In two dimensions the distance is the square root of (x2 − x1) squared plus (y2 − y1) squared. In three dimensions one more squared difference joins the sum under the root: the square root of (x2 − x1) squared plus (y2 − y1) squared plus (z2 − z1) squared. That is not a new idea pasted on top of the old one; it is the same Pythagoras argument applied twice, first across the floor and then up to the height. A point in space is first projected down onto the floor, the floor distance is found, and then the floor distance and the height difference form the two legs of a second right triangle whose hypotenuse is the true distance. Every extra dimension adds one more squared term, and nothing else changes.

Check the second row by hand, because it is the example the FAQ uses and it is worth seeing once in full. The x-coordinates 1 and 5 average to 3, the y-coordinates 2 and 6 average to 4, and the z-coordinates 3 and 11 average to 7, so the midpoint is (3, 4, 7). The differences are 4, 4 and 8; their squares are 16, 16 and 64; the sum is 96; and the square root of 96 simplifies to 4√6, about 9.80. The third row shows why negative coordinates need no special treatment: −3 and 5 average to 1, 1 and −5 average to −2, 4 and 2 average to 3, and the differences 8, −6 and −2 square to 64, 36 and 4, whose sum is 104. Squaring removes the signs, which is exactly as it should be, because a distance does not care which direction you travelled. The fourth row is the page default with z set to 0, included so you can see the two-dimensional answer sitting inside the three-dimensional one rather than beside it.

Three-dimensional midpoints turn up wherever a point has a height as well as a position: the centre of a room, the halfway point of a flight path, the middle of a beam in a structure, the centre of a sphere when two ends of a diameter are known. The last of these is a neat reuse of the tool: the centre of a sphere is the midpoint of any diameter, so two opposite points on a sphere and this calculator are all you need to find its centre, and half the distance between them is its radius. The same trick already appears in two dimensions for circles, where the centre is the midpoint of any diameter, and the third coordinate changes nothing about the reasoning. If your problem gives you coordinates with a z value, use the Advanced panel rather than dropping the z and hoping: dropping it silently replaces your segment with its shadow on the floor, and the midpoint of the shadow is only correct when both points share the same height.

Dividing a segment in a ratio: the section formula

The midpoint is a special case of a wider question: not the halfway point, but the point that sits a given fraction of the way along a segment. Dividing the segment from A to B in the ratio m:n, measured from A, means the point is m parts along a segment cut into m + n parts, so it sits m ÷ (m + n) of the way from A to B. The section formula says exactly that, coordinate by coordinate: the point is ((n × x1 + m × x2) ÷ (m + n), (n × y1 + m × y2) ÷ (m + n)). The weighting looks backwards at first glance — the x1 term is multiplied by n, the far part of the ratio — but it is correct, and the midpoint proves it: with m = n = 1, each coordinate becomes (x1 + x2) ÷ 2, the familiar average. The weight on each endpoint is the share of the segment on the other side of the point, which is why a point close to A is weighted heavily toward A.

Work the default segment through the formula once, with the ratio 1:2, the Advanced panel default. The segment runs from A (2, 3) to B (6, 9), and the point dividing it in the ratio 1:2 from A is ((2 × 2 + 1 × 6) ÷ 3, (2 × 3 + 1 × 9) ÷ 3), which is (10 ÷ 3, 15 ÷ 3), or about (3.33, 5). It sits one third of the way from A to B, closer to A, as a 1:2 split should be: the distance from A to the point is one third of the whole distance, and the distance from the point to B is the remaining two thirds. The Advanced panel computes this point live from whatever points and ratio you enter, so you can watch it slide along the segment as the ratio changes — push m up and the point walks toward B, push n up and it retreats toward A, and at 1:1 it lands exactly on the midpoint the main calculator reports.

Every entry in that table is the same formula with different weights, and the fraction column is the check that keeps the weighting honest. For 2:1, the point is two thirds of the way along, so x is (1 × 2 + 2 × 6) ÷ 3 = 14 ÷ 3, about 4.67, and y is (1 × 3 + 2 × 9) ÷ 3 = 21 ÷ 3 = 7. For 1:3, the point is one quarter of the way along, so x is (3 × 2 + 1 × 6) ÷ 4 = 3 and y is (3 × 3 + 1 × 9) ÷ 4 = 4.5. The symmetry is worth noticing: the 1:2 and 2:1 points are mirror images in the segment, equidistant from the midpoint, and the same holds for 1:3 and 3:1. If your two ratio answers do not mirror like that, one of them has its weights swapped, which is the single most common section-formula error.

The same formula divides a segment from the outside as well, though the sign changes. A point dividing AB externally in the ratio m:n — beyond B, so that its distances to A and B are in the ratio m:n — is ((m × x2 − n × x1) ÷ (m − n), (m × y2 − n × y1) ÷ (m − n)). On the default segment, the external 2:1 point is ((2 × 6 − 1 × 2) ÷ 1, (2 × 9 − 1 × 3) ÷ 1) = (10, 15): beyond B, twice as far from A as from B, since the distance from (10, 15) to A is √208 and to B is √52, a ratio of exactly 2:1. External division needs m and n to differ — at 1:1 the denominator is zero, which is the algebra telling you that no finite point is equally far from A and B while lying beyond the segment. The Advanced panel on this page covers internal division, the case homework and design work ask for; external division is included here in the guide so the formula is on record, with its worked check, rather than left as a gap.

Section division is also how several famous points are defined. The centroid of a triangle — the point where its three medians meet — divides each median in the ratio 2:1, measured from the vertex, which is why the centroid is the average of the three vertices: each median runs from a vertex to the midpoint of the opposite side, and the 2:1 point along it lands on the same spot whichever median you start from. Trisection points, which cut a segment into three equal parts, are the 1:2 and 2:1 points of the table above. And in design and mapping work, placing a label or a stop a stated fraction along a route is a section-formula calculation whether or not anyone names it: a point 40 percent along a road from one junction to another is the 2:3 point of that segment, found by exactly the weighting this section describes.

Reading the plot: what the graph is already telling you

The plot above the guide is not decoration; it is the same calculation drawn, and it is worth learning to read it as a check on the numbers. The two dots are your endpoints, the dashed line is the segment joining them, and the diamond is the midpoint. The plot redraws itself with every keystroke and rescales to fit whatever points you enter, keeping equal spacing on both axes, so the diamond staying exactly halfway along the dashed line is a visual proof of the arithmetic, not an illustration of it. If you ever mistype a coordinate, the plot is usually where you notice: a point landing in the wrong quadrant, or a segment far longer than you expected, is easier to spot as a picture than as a pair of numbers.

Two features of the plot deserve a deliberate look. First, the midpoint splits the segment into two equal halves, and the note under the plot states the half-length: for the default points the halves are 3.61 units each, which is half of the 7.21 distance. That equality is the definition of a midpoint, and it is a stronger check than the coordinates alone, because a point can sit on the segment without sitting halfway along it — the section points of the previous section all lie on the same dashed line, but only the midpoint splits it equally. Second, the plot keeps its grid square: one unit horizontally is the same length as one unit vertically. A plot that stretched one axis would still place the diamond correctly by coordinate, but the segment would look tilted wrongly and the equal halves would stop looking equal, which is why the rescaling preserves equal spacing and changes the grid step instead when your coordinates grow large.

The plot also settles a question the numbers alone leave abstract: what happens at the edges. Enter the same point twice and the segment collapses to a single dot, the midpoint is the point itself, and the distance is 0 — the plot shows one mark doing three jobs. Enter two points sharing an x value and the segment stands vertical, the slope undefined, but the midpoint is as ordinary as ever, because averaging two identical x values returns that same x value. Enter points far apart and the grid step coarsens so everything still fits, with the plot note saying so; the diamond does not move off the halfway mark when the scale changes, because scaling a picture cannot move a halfway point. Those edge cases are all handled by the same averaging, and the plot is the quickest way to convince yourself of that.

Worked examples beyond the default

The default example (2, 3) to (6, 9) is deliberately friendly — whole numbers, a midpoint of (4, 6), a distance of √52. Real work is less tidy, so here are three further examples worked in full, including the two cases that most often cause slips: negative coordinates, and coordinates whose average lands on a half.

First, (−4, 5) to (6, −3). The x values −4 and 6 add to 2 and halve to 1; the y values 5 and −3 add to 2 and halve to 1, so the midpoint is (1, 1) — a tidy answer hiding a crossing of both axes, since the segment passes from the second quadrant to the fourth. The differences are 10 and −8, whose squares are 100 and 64, summing to 164, whose root simplifies to 2√41, about 12.81. The slope is the rise −8 over the run 10, which is −0.8: the line falls as it goes right, which matches a segment whose second point is lower than its first. Second, (−3, 5) to (5, 5): a horizontal segment, so the y values are already equal and the midpoint keeps that y, (1, 5); the distance is simply the run, 8, and the slope is 0. Third, (1, 2) to (4, 7): the x values add to 5 and halve to 2.5, the y values add to 9 and halve to 4.5, so the midpoint is (2.5, 4.5) — halves are exact, not approximations, and this page writes them as fractions where it helps, 5/2 and 9/2. The differences are 3 and 5, squares 9 and 25, sum 34, root about 5.83, and the slope is 5 over 3, about 1.67.

Two habits make these examples reliable. One: average the coordinates before doing anything else — do not find the distance or the slope first and then try to work backwards, because the midpoint needs only the averages and every extra step is a chance to carry a wrong number forward. Two: check the midpoint by distance, not by re-averaging. The midpoint must be the same distance from both endpoints, and that distance must be half the total: for the first example, the distance from (1, 1) to either endpoint is √41, about 6.40, which is exactly half of 12.81. If your candidate midpoint is not equidistant from the two endpoints, it is not the midpoint, however plausible its coordinates look. That check catches the most common slip, which is averaging x with y — adding the wrong pair — because the resulting point almost never lands equidistant.

Mistakes that move the midpoint

Most midpoint errors are small slips with large consequences, and they cluster into four kinds. The first is averaging the wrong pair: adding x1 to y2, or averaging all four numbers into a single value, which produces a number rather than a point and cannot be plotted. The guard is dimensional: a midpoint has two coordinates, each averaged from its own kind, x with x and y with y. The second is halving only once: adding the coordinates and forgetting to divide by two gives the sum point, which is twice as far from the origin as the midpoint along each axis — on the default example that would be (8, 12) instead of (4, 6), a point beyond B itself. If your answer lies outside the segment, you have almost certainly forgotten the halving, because a midpoint always lies between its endpoints.

The third kind is a sign slip, and it loves negative coordinates: averaging −4 and 6 is 1, not −5 or 5, because the sum is 2 before halving, and the sign of the sum decides the sign of the answer. Working with the sum written out — (−4 + 6) ÷ 2 — rather than doing it in your head is the cheapest insurance available. The fourth kind is confusing the midpoint with the distance or the slope, usually by reporting a single number where a point belongs: 7.21 is the distance between the default points, 1.5 is their slope, and neither is the midpoint (4, 6). The three answers describe different things — a place, a length and a tilt — and the calculator reports all three side by side precisely so they can be told apart. When an answer is asked for as a point, it must arrive as a pair in parentheses; a bare number is the wrong kind of thing, even if it was computed correctly as something else.

One further confusion is worth naming because it appears in classrooms every year: the midpoint of a segment is not the same as the average of the distances, and it is not the point halfway along the axes separately in any other sense than the formula states. It is also not, in general, the point where the segment crosses an axis — the default segment crosses neither axis between its endpoints, and its midpoint (4, 6) is nowhere near an axis. Axis crossings are found by setting a coordinate to zero and solving, which is a different calculation with a different answer. Keeping the three questions separate — where is halfway, how long is it, how steep is it, where does it cross — is most of the subject, and the calculator’s three tiles exist to keep them separate on screen as well as in thought.

Where midpoints do real work

Beyond homework, midpoints are load-bearing in several trades, usually under other names. In computer graphics, the midpoint of a segment is the first step of subdivision: splitting a line or a curve at its midpoint, again and again, is how smooth curves are drawn and how meshes are refined, and the midpoint formula is the operation repeated at every step. In mapping and surveying, the midpoint of a boundary segment is a natural place for a marker or a label, and dividing a route in a stated ratio — the section formula of this page — places waypoints at measured fractions of a journey. In construction and carpentry, finding the centre of a span, a board or a room is a midpoint calculation, often done by measuring from both ends and meeting in the middle, which is the physical version of averaging the two endpoint coordinates.

In statistics and data work, the midpoint of a class interval — the average of its lower and upper bounds — stands in for the whole class when a mean is estimated from grouped data, and the midpoint of a data range is a quick centre before any formal measure is computed. In physics, the centre of mass of two equal masses sits at the midpoint of the segment joining them, and for unequal masses it sits at the section point weighted toward the heavier mass, which is the section formula with physical weights instead of ratio parts. None of these uses needs a new formula; they all need the same averaging, applied to coordinates, measurements or masses as the problem demands. That reusability is why the formula is taught early and used late: it is one of the few pieces of school geometry that never gets replaced, only generalised — to three dimensions, to ratios, and to weighted averages, all of which this page now covers.

Beyond halfway: the section formula

The midpoint divides a segment 1:1. The section formula generalises it: the point dividing (x1, y1) to (x2, y2) in the ratio m:n is ((m times x2 plus n times x1) divided by (m plus n), similarly for y). At m equals n equals 1 it reduces to the midpoint formula. The companion guide works a 1:2 example in full.

Midpoint, centroid, and the average idea

The midpoint averages two points; the centroid of a triangle averages three vertices. The pattern generalises: the average of any set of points is their centre, and the midpoint is the two-point case. Seeing the pattern makes both formulas harder to forget.

Related guides

The wider coordinate toolkit

Everything that shares this page’s two points, one click away.