GD&T Reference: Complete Guide

What is GD&T?

Geometric Dimensioning and Tolerancing (GD&T) is a symbolic language used on engineering drawings and models to define the allowable variation in the geometry of manufactured parts. It communicates not just size but shape, orientation, location, and runout requirements in a way that corresponds directly to the part's function in an assembly.

Unlike coordinate tolerancing—where a dimension is given a simple ± value—GD&T specifies a geometric tolerance zone whose shape and size are tied to what the feature must actually do. A hole that must locate a pin, for instance, gets a cylindrical tolerance zone for its axis rather than two independent linear bands. This is a more complete and usually more forgiving description of acceptable variation.

GD&T originated from work done by Stanley Parker at the Royal Torpedo Factory in the 1940s and was formalised into industry standards that most engineering organisations now follow in some form.

Why GD&T matters in manufacturing

From a production standpoint, GD&T matters for three reasons: it reduces ambiguity, it typically increases the usable tolerance zone compared to equivalent coordinate tolerancing, and it provides a clear basis for inspection.

Ambiguity is costly. When a drawing can be interpreted in more than one way, the machinist, the CMM programmer, and the quality engineer may each use a different interpretation. Parts get rejected that are functional, or—worse—accepted that are not. GD&T, when applied correctly to the governing standard, eliminates most of that ambiguity by fully defining the tolerance zone shape and the reference frame used to measure it.

The inspection benefit is equally practical. A well-written GD&T callout tells the inspection team exactly which datums to establish, in what order, and what zone the feature must fall within. This means CMM programmes can be written directly from the drawing without engineering interpretation at the machine.

Key standards: ASME Y14.5 and ISO GPS

Two standards dominate global GD&T practice. ASME Y14.5 (current edition: 2018) is the standard most commonly used in North America. ISO GPS (Geometrical Product Specifications) is the framework used in Europe and increasingly across global supply chains; it is not a single document but a family of interconnected ISO standards.

The two systems share the same foundational concepts—datums, feature control frames, symbolic notation—but they differ in important details. ASME Y14.5 applies Rule #1 (the envelope requirement) by default to features of size, meaning the feature must not violate its maximum material boundary unless otherwise specified. ISO GPS does not apply an equivalent rule by default; perfect form at MMC must be explicitly invoked with the circle-E modifier. Modifier symbols, datum target notation, and the handling of compound tolerances also differ between the two.

Always specify the governing standard in the drawing title block. Mixing rules from both standards on the same drawing creates the exact ambiguity GD&T is meant to eliminate.

GD&T symbols overview

GD&T uses fourteen geometric characteristic symbols grouped into five categories:

• Form: Straightness, Flatness, Circularity, Cylindricity
• Profile: Profile of a Line, Profile of a Surface
• Orientation: Angularity, Perpendicularity, Parallelism
• Location: Position, Concentricity, Symmetry (note: Concentricity and Symmetry were removed in ASME Y14.5-2018; use Position or surface profile controls instead)
• Runout: Circular Runout, Total Runout

Each symbol appears in a feature control frame alongside the tolerance value and, where applicable, material condition modifiers (⌀, M, L, S) and datum references. Reading a feature control frame correctly—left to right, understanding each compartment—is a foundational skill before applying any of the individual controls.

Datums and datum reference frames

A datum is a theoretically exact geometric reference—a plane, axis, or point—derived from a physical datum feature on the part. Datum features are actual surfaces or features of size that the part rests against or is constrained by during measurement, replicating how it is held in its assembly fixture.

The datum reference frame (DRF) is the three-plane system constructed from the datum features in a defined sequence: primary, secondary, and tertiary. The primary datum constrains the most degrees of freedom (typically three for a flat surface), the secondary constrains additional degrees, and the tertiary removes the final rotational or translational freedom.

The sequence matters enormously. If you reverse the primary and secondary datums, you change which surface the part is seated against, which changes where the part is located in space during measurement, and therefore which features pass or fail. The datum sequence must reflect functional assembly priority—the surface that contacts the mating part first should be the primary datum.

Tolerance types and their controls

Form controls govern the shape of a single feature in isolation—they do not reference any datum. Flatness on a surface, cylindricity on a bore. These are the most self-contained controls and are often applied to establish the datum features themselves.

Orientation controls (angularity, perpendicularity, parallelism) govern the angle of a feature relative to a datum. They always require at least one datum reference in the feature control frame.

Location controls govern where a feature is relative to a datum reference frame. Position is the most widely used location control and is the primary tool for controlling hole locations, slot centrelines, and coaxial features. Position tolerancing can interact with the size of a feature of size through material condition modifiers, which introduces bonus tolerance—additional positional tolerance available as the feature departs from its maximum material condition.

Profile controls are the most versatile controls in the system. Profile of a surface can simultaneously control form, orientation, and location of complex surfaces with a single callout, and in ASME Y14.5-2018, unequal profile distribution is explicitly notated.

Runout controls are used specifically for rotating parts. Circular runout measures the variation at each cross-section as the part is rotated about its datum axis; total runout accumulates variation across the entire surface.

Common mistakes on the shop floor

After fourteen years managing production and quality across machined and fabricated components, these are the errors I see repeated most often:

• Applying GD&T without specifying the governing standard. The drawing says "GD&T per…" and then the field is blank. Inspection then defaults to whatever standard the CMM software assumes.
• Selecting datums based on drawing convenience rather than function. Datum features should be the surfaces that locate the part in its assembly fixture, not whatever edge is easiest to dimension from.
• Confusing the tolerance value with the tolerance zone diameter. A position tolerance of ⌀0.2 means the axis must fall within a cylinder of 0.2 mm diameter—not ±0.2 mm in each direction, which would be a much larger zone.
• Ignoring material condition modifiers. Leaving a position callout at RFS (regardless of feature size) when an MMC modifier would both reflect functional intent and relax the tolerance on oversized holes.
• Over-tolerancing form controls. Applying a flatness tolerance tighter than the surface finish achievable by the specified process, or tighter than necessary for the sealing function intended.

Going deeper: linked articles in this cluster

This hub page provides the framework. The articles below go into the calculation detail and practical application for specific controls.

For location tolerancing, understanding how to convert CMM coordinate output into a reportable position error is essential. Our article on True Position Calculation: Formula & Worked Example covers the diametrical position error formula step by step, with a complete numerical example you can follow alongside your own inspection data.

Once you are comfortable with the basic position calculation, the next layer is material condition modifiers. When an MMC modifier is applied to a position tolerance, the permissible positional error increases as the feature departs from its maximum material condition—this is bonus tolerance. Our article on Position Tolerance with MMC: Bonus Tolerance explains how bonus tolerance is calculated, when it is appropriate to invoke it, and how it affects your acceptance criteria at inspection.

Further articles in this cluster will address profile tolerancing, datum target application, and tolerance stack-up analysis. Each will link back to this reference page and to one another where controls interact.

Frequently asked questions

What is the difference between GD&T and traditional plus/minus tolerancing?

Plus/minus tolerancing controls each dimension independently along a single axis. GD&T uses geometric controls that describe the actual function of a feature—position, form, orientation—often creating a tolerance zone that better matches how the part works in an assembly. This typically allows more usable tolerance while maintaining functional fit.

Which standard should I follow—ASME Y14.5 or ISO GPS?

That depends on where your product is designed and manufactured. US-based programmes generally follow ASME Y14.5-2018. European and globally distributed programmes more commonly use the ISO GPS framework. The two are not interchangeable; default rules for modifiers, datum precedence, and some symbol definitions differ. Specify the governing standard in your drawing title block.

What is a datum and why does datum sequence matter?

A datum is a theoretically exact point, axis, or plane derived from a physical datum feature on the part. The sequence—primary, secondary, tertiary—determines which degree of freedom each datum constrains. Changing the order changes which surfaces locate the part and can shift the entire tolerance analysis, so the sequence must reflect functional assembly priority.

Can bonus tolerance be applied to all GD&T controls?

No. Bonus tolerance is only available when the Maximum Material Condition (MMC) or Least Material Condition (LMC) modifier is applied to a feature of size. It cannot be applied to form controls such as flatness or straightness, because those controls reference the feature itself rather than a datum, and MMC/LMC modifiers are not permitted on them.

How do I calculate true position?

True position is calculated from the measured deviation of a feature's actual axis or centre plane from its theoretically exact location. The formula uses the diametrical position error: TP = 2 × √(Δx² + Δy²) for a two-axis case. A full explanation with a worked example is available in our true position calculation article.