How to Calculate True Position in GD&T

September 19, 2026
Category: Blog

True position is one of the most miscalculated controls in GD&T, not because the math is hard, but because the drawing is easy to misread.

I have watched skilled engineers get the arithmetic right and still land on the wrong answer. The problem almost never lives in the square root.

It hides in the details. People pull coordinates from the wrong dimensions. They swap the datum sequence. They forget to multiply by 2. They add MMC bonus tolerance when the callout never allowed it.

Each mistake looks small. Each one flips a good part into a reject, or lets a bad part ship.

This guide focuses on the parts that actually decide your result: how to read the callout, how to run the calculation, and how to avoid the inspection errors that trip up so many people.

What True Position Actually Controls

Before any formula, get one idea straight. True position controls where a feature is allowed to be, not how big that feature is.

That single distinction clears up more confusion than any definition ever will.

It Controls Location

Position pins down the location of a feature — usually the axis of a hole or shaft, or the center plane of a slot. It asks a simple question: did that feature land where the drawing said it should?

It Does Not Control Size or Form

Position says nothing about how big the hole is, how smooth the surface finished, or whether the feature holds its shape on its own. Those are separate checks with separate tolerances. A hole can meet its size tolerance and still fail position if its axis falls outside the allowed zone.

The Zone Is Diametrical

Here’s the part people miss. The tolerance zone is diametrical — a circle in 2D thinking, a cylinder in 3D. That’s why the calculated result gets compared to a diameter value, not a simple X or Y offset.

Picture it this way: a hole may still measure Ø10.00 and pass its size check, but if its center shifts too far from basic location, it fails true position. Right size, wrong place. If location drifts far enough, parts stop assembling, no matter how perfect the diameters look.

Read the Drawing First: What You Need Before You Calculate

Before you calculate anything, read the feature control frame as a set of instructions, not just a number.

That frame tells you what is being controlled, how large the tolerance zone is, whether bonus tolerance is allowed, and which datum structure defines the measurement frame. Rush past it and every number that follows carries your mistake forward.

Break the Feature Control Frame Apart

Don’t start calculating the moment you spot a number. Confirm four things first:

The geometric control is position.
The tolerance is diametrical.
The modifier is RFS, MMC, or LMC.
The datum references, and their order.

Pull Coordinates from Basic Dimensions Only

The coordinates you use in the formula come from basic dimensions, not from general drawing dimensions. Those boxed numbers define the exact intended location and carry no plus/minus of their own.

Grab the wrong nominal values and the rest of the calculation can be mathematically perfect and still wrong.

Respect the Datum Order

Datum order is not formatting. It defines how the part is oriented and how location is measured. Primary, secondary, and tertiary datums build the reference frame in that exact sequence.

Swap the order and you’ve built a different coordinate system. The same measured point can pass in one frame and fail in another. If the datum reference frame is wrong, the result is meaningless even when the formula is right.

Read the Modifier for Its Effect

You don’t need a full lecture on material conditions here. You just need to know how each one touches the allowable tolerance:

RFS: no bonus.

MMC: bonus possible, based on actual feature size.

LMC: different logic, used in specific cases.

No modifier shown: assume RFS unless the drawing standard says otherwise.

The modifier changes the allowable tolerance, not the way the base formula runs. Most confusion starts when people mix up the measured true position value with the allowed tolerance after MMC bonus.

Drawing element	What to identify	Why it matters
Position symbol	Confirms positional control	Keeps it separate from size or profile
Diameter symbol	Shows the diametrical zone	Explains why the result compares to a diameter
MMC/LMC/RFS	Affects allowable tolerance	Changes pass/fail logic
Datum sequence	Establishes the reference frame	Wrong order gives wrong result
Basic dimensions	Gives nominal coordinates	Inputs for the calculation

This kind of print discipline is exactly what separates a clean CNC machining service from one that ships surprises.

The True Position Formula, Explained Simply

Here’s the formula for a 2D check:

TP = 2 × √(dX² + dY²)

The formula itself is simple. The important part is understanding what kind of value it produces.

What dX and dY Are

They’re the deviations: measured minus basic, for each axis. They can come out positive or negative. Direction matters while you subtract, but it drops out once you square the values.

Why You Square and Add

Position error isn’t a one-axis problem. A feature can drift in both X and Y at the same time. Squaring each deviation and adding them captures that combined offset in space, then the square root pulls it back to a single distance.

Why You Multiply by 2

This is the step people forget, so pay attention here.

The square root term gives the radial deviation from true location — a radius. Position tolerance is diametrical. Multiplying by two converts that radial offset into the diametrical value your tolerance actually uses.

Don’t skip this: skipping the final multiply-by-two step can make a bad part look acceptable. You’d be reporting half the real deviation.

The same logic extends to 3D by adding dZ under the root. But only reach for 3D when the drawing and measurement setup genuinely call for it.

How to Calculate True Position Step by Step

A true position check becomes much easier when you separate the measured result from the allowable tolerance.

First calculate what the feature actually did. Then decide how much deviation the drawing allows. Those two numbers are related, but they are not the same thing.

Here’s the flow, one action per step:

1. Identify the basic location. Pull X and Y from the boxed dimensions and confirm your units.

2. Record the measured location. Take the actual coordinates from a CMM, vision system, or manual setup.

3. Calculate dX and dY. Measured X minus basic X. Measured Y minus basic Y.

4. Square each deviation. This removes the sign and captures magnitude.

5. Add them and take the square root. That gives the radial offset.

6. Multiply by 2. Now you have the diametrical true position value.

7. Compare to the stated tolerance. Under RFS, compare directly. Under MMC or LMC, adjust the allowable tolerance first.

8. Make the call. Report clearly, and never mix the measured value with the allowable value.

Step	Action	Example value
1	Basic X, Y	2.000, 1.500
2	Measured X, Y	2.008, 1.494
3	dX, dY	0.008, −0.006
4	Square deviations	0.000064, 0.000036
5	Sum and square root	0.010
6	Multiply by 2	0.020
	7	Compare to tolerance	0.020 vs 0.030
8	Result	Pass

Worked Example: Pass or Fail Under RFS

Let me walk you through the cleanest version of the check. One hole, two dimensions, no modifiers. Every harder case builds on this.

Under RFS, the decision is straightforward: calculate the true position value and compare it directly to the stated tolerance.

The Setup

Picture a single hole on a bracket. Basic location: X 2.000, Y 1.500. The feature control frame calls for a position tolerance of 0.030, diametrical, with no modifier attached.

That missing modifier matters. No modifier defaults to RFS, so no bonus tolerance will shift the limit. The pass/fail decision rides entirely on the measured positional error.

The Math

The part comes off the machine measuring X 2.008, Y 1.494.

dX = 2.008 − 2.000 = 0.008
dY = 1.494 − 1.500 = −0.006

Square and add:

0.008² = 0.000064
0.006² = 0.000036
Sum = 0.000100

Square root: √0.000100 = 0.010. That’s the radial offset. Multiply by 2 for the diametrical value:

TP = 2 × 0.010 = 0.020

The Call

The result is 0.020. The tolerance is 0.030. It sits inside the limit, so the hole passes.

Notice how clean the judgment was. One value, one limit, one decision. Nothing moved the allowable side, because RFS holds the limit fixed regardless of hole size. Burn this rhythm in — find the deviations, combine them, double the result, compare — and the harder cases stop feeling new.

Worked Example: How Bonus Tolerance Works at MMC

Here’s the idea that clears up most MMC confusion: the position formula does not change. MMC changes how much positional error the drawing allows, not how you calculate it.

Same Hole, One New Detail

Reuse that bracket hole, but add a size tolerance. The drawing now calls out a position tolerance of 0.030 at MMC, with an MMC hole size of 10.000 — the smallest allowed diameter.

When the hole runs bigger than MMC, it clears its mating pin more easily. That extra clearance buys extra positional freedom. We call it bonus tolerance.

Working the Numbers

Inspection reports the hole at 10.012.

Actual size − MMC size = bonus
10.012 − 10.000 = 0.012

Add that bonus to the stated tolerance — to the allowable limit, not the measured result:

Total allowable = 0.030 + 0.012 = 0.042

Say the measured true position comes out to 0.038, calculated the same way as always. Compare it to the new limit:

Item	Value
Stated position tolerance at MMC	0.030
MMC hole size	10.000
Actual hole size	10.012
Bonus tolerance	0.012
Total allowable position	0.042
Measured TP result	0.038
Outcome	Pass

A 0.038 result would have failed under a plain 0.030 tolerance. The larger hole created enough bonus to accept it. Same error, different outcome, driven entirely by size.

The Mistake I See Most

People add bonus to the wrong number. They calculate 0.038, then adjust that measured value by the bonus and report a made-up figure. That’s backwards.

Keep the order clean: the measured result stands on its own. Bonus belongs on the allowable side. Calculate the true position, calculate the bonus from actual size, grow the stated tolerance, then compare.

A Quick Word on LMC

LMC flips the logic — bonus grows as the feature moves toward its largest material state, often to protect a minimum wall thickness. It behaves differently enough that you should never apply MMC bonus rules to an LMC callout.

Pattern of Holes: What Changes and What Does Not

Move from one hole to a bolt circle and part of you wants to treat it as a new problem. It isn’t. But it isn’t a straight copy-paste either.

What Stays the Same

Every hole still has a center, a basic location, and a measured location. You run the same square-root-then-double routine per feature. Your calculation habits don’t go out the window.

What Changes

A pattern callout often controls the holes as a group. A hole pattern still involves individual feature locations, but the drawing may control the group in ways a single-hole calculation never captures. Spacing between holes and the rotation of the whole set can matter as much as any one center. A pattern can twist as a unit while each hole looks fine alone.

Composite Callouts Are Different

A composite callout looks like a stacked, two-line version of a normal one. It is not read the same way. The upper segment usually controls the pattern’s location to the datums; the lower segment controls the tighter feature-to-feature relationship. If the callout is composite, don’t assume simple single-segment logic applies. For dense patterns, inspection software or a formal standards review earns its keep.

How to Check True Position Without a CMM

Good news up front: you do not need a CMM to calculate true position, but you do need trustworthy coordinates.

The formula never asked for a coordinate measuring machine. It asked for where the feature should be and where it landed.

Where Coordinates Come From

Height gauges — simple layouts off clean reference surfaces.

Optical comparators — smaller flat parts measured against known references.

Vision systems — fast coordinate capture on flat features.

Manual setups — surface plates and indicators, when built with care.

Functional gauges — go/no-go checks that confirm fit without a number.

The Catch

The real issue is not the formula. It is whether your method can build the same datum reference frame the drawing requires. A CMM builds that frame in software. Manual methods lean on your fixturing, and every clamp and alignment adds a little slop.

That’s also why a hand calculation and a CMM result won’t always match — different feature extraction, different datum simulation, different rounding.

A functional gauge answers a different question entirely: not what the true position value is, but whether the part clears a functional boundary. Clean pass/fail, no number.

Method	Best for	Main limitation
CMM	Complex geometry, tight tolerances	Cost and programming time
Vision system	Flat parts, fast coordinate checks	Setup and lighting sensitivity
Height gauge / manual	Simple layouts, looser tolerances	Higher operator and alignment error
Functional gauge	Production pass/fail checks	Gives no numeric TP result

For simple flat features with generous tolerances, the bench holds up fine. Push toward tight tolerances or angled axes and confidence drains fast — that’s when the CMM earns its place. Reliable location data is the backbone of solid inspection and shipping, whatever tool produces it.

Common Mistakes That Cause Bad Results

Most true position mistakes do not come from difficult math. They come from wrong inputs or a comparison against the wrong limit.

Forgetting to multiply by 2

Reports half the deviation. Always compare a diametrical result to a diametrical tolerance.

Wrong datum reference frame

Builds a different coordinate system. Rebuild orientation straight from the feature control frame.

Pulling non-basic dimensions

Wrong nominal target. Use basic dimensions only.

Adding MMC bonus with no modifier

Inflates the limit. No modifier means no bonus.

Comparing to size tolerance

Wrong pass/fail basis. Compare to position tolerance.

Rounding too early

Distorts borderline calls. Round only at final reporting.

Mixing 2D and 3D logic

Zone geometry no longer matches the callout. Match the math to the setup.

Measuring the wrong feature representation

Position controls a derived axis or center plane, not a surface point. Extract the true feature.

Notice what these share. Not one is a math failure. Slow down at the front end, and most disappear before you touch the calculator. That same discipline underpins reliable CNC metal machining across every material.

Quick Checks Before You Sign Off a Result

A short sanity check takes less time than sorting out a false reject later. Run these seven before your name goes on the report:

✓Units match the drawing.

✓Basic dimensions were used.

✓Datum order matches the callout.

✓The result went against the position tolerance, not the size tolerance.

✓Bonus tolerance applied only when a modifier allowed it.

✓The number makes geometric sense against what you saw on the part.

✓Software and manual setups share the same reference assumptions.

In practice, these last checks catch more errors than redoing the square root.

True Position vs Other Controls: Keep It Short

Not every location-related problem is a true position problem.

Position is usually the right control for holes, pins, and repeatable feature locations. It fences off a diametrical zone and asks whether the axis or center plane landed inside it. Profile and concentricity answer different questions.

Control	Main purpose	Typical use
Position	Controls feature location	Holes, shafts, bolt patterns
Profile	Controls form and location together	Complex surfaces and contours
Concentricity	Controls derived median points	Limited, specialized cases
Symmetry	Controls median-plane relationship	Specific center-plane cases

When a callout makes you pause, ask what it’s trying to protect. Where a feature sits? Position. The shape of a surface along with its place? Profile. How one derived center relates to another? Now you’re in concentricity or symmetry, and you should confirm the part really needs that level of control.

Frequently Asked Questions

Q: What is the true position formula?

A: For 2D: TP = 2 × √(dX² + dY²). For 3D, add dZ under the root. Find the deviations, combine them, double the result.

Q: Why do you multiply by 2?

A: The square root gives a radius. Position tolerance is diametrical. Multiplying by 2 converts the radial offset into a diameter so you’re comparing like with like.

Q: How do you calculate true position at MMC?

A: Exactly the same way. Only the allowable limit changes. Add the bonus from actual size to the stated tolerance, then compare your unchanged result against the larger limit.

Q: Can true position be checked without a CMM?

A: Yes, with reliable coordinates and a datum frame you can trust. Height gauges, vision systems, and careful manual setups all work. Functional gauges give pass/fail without a number.

Q: Why is my hand calculation different from my CMM result?

A: Different feature extraction, different datum simulation, early rounding, or a 2D-versus-3D mismatch. Neither is wrong on purpose; they built the answer differently.

Q: Does true position apply to slots?

A: Yes. For a hole you locate an axis; for a slot you usually locate a center plane, and the zone shifts to parallel planes to match.

Q: Is a lower true position value always better?

A: Generally yes, since it means closer to nominal. But don’t chase zero. If the part passes comfortably, it functions as intended.

Q: What’s the difference between true position and coordinate tolerance?

A: Zone shape. Coordinate tolerance creates a rectangular zone from plus/minus limits. True position uses a round or cylindrical zone that treats deviation the same in every direction, which protects the assembly more honestly.

Where the Real Skill Lives

After all the math, here’s the truth I keep coming back to: true position was never the hard part. The formula behaves. It gives an honest number every time you feed it honest inputs.

Separate three things clearly and it all settles down: the nominal location, the measured deviation, and the allowable tolerance. The nominal comes from your basic dimensions. The deviation comes from what the feature did. The allowable comes from the tolerance and, when the callout permits it, the bonus.

That’s why reading the drawing has to come first. The formula only rewards you when the nominal location, the datum structure, and the modifier were all read correctly before you started. Correct interpretation is what keeps a clean calculation from becoming a bad decision.

Hold on to the one idea that clears up modifiers: MMC changes the allowable tolerance, nothing else. Calculate the same way you always do, then let actual feature size adjust how much room the drawing gives you.

So if I leave you with anything, let it be this. You don’t win at true position by memorizing more definitions. You win with discipline — confirming your inputs, respecting the datum frame, applying bonus only when the callout allows it, and checking against the right limit. Do that consistently, and the mistakes that plague so many inspections quietly stop happening. The math was always going to cooperate. The rest is on you.

James Cao CNC machining expert

We specialize in providing CNC milling, CNC turning, machinery manufacturing, and sheet metal fabrication services.

Share Blog:

How to Calculate True Position in GD&T