Even the most well-intentioned engineering drawings can fall victim to geometric dimensioning and tolerancing(GD&T) errors that ripple across the entire manufacturing process. From confusion on the shop floor to rejected parts and production delays, small mistakes in GD&T can have big consequences. While GD&T is designed to provide clarity and control, misusing its principles often leads to the exact opposite: uncertainty, inconsistency, and increased costs.

This article highlights some of the most common missteps in applying GD&T and datum features, mistakes that frustrate machinists, puzzle quality inspectors, and derail CMM programmers. Whether you’re new to GD&T or a seasoned pro, identifying these issues early can improve communication, streamline inspection, and ultimately result in better parts.

Let’s start with improper Datums. Datum features are the backbone of any GD&T scheme, but they’re often misapplied or underdefined. The most common errors include:

• Choosing weak or unstable datum features: Surfaces that wobble, rock, or don’t repeat consistently from part to part make poor datum selections. Yet designers often select these for convenience, rather than functional intent.
• Missing functional relationships: A datum should represent how the part is mounted, assembled, or functions in the real world. When datum features are chosen simply for manufacturability or based on ease of measurement, the design intent is lost.
• Failing to define a complete datum reference frame: Often only a primary datum is called out, leaving the part free to rotate or shift in ways that make inspection impossible or ambiguous. Now in some cases, leaving a degree or two of freedom unrestrained is part of design intent and function.

Fix: Ensure datum features reflect functional requirements and clearly define a complete 3-plane reference frame when needed. Collaborate across design, manufacturing, and quality teams to align on what matters most. I am very proud of companies when I see in the Title Block a clear system of collaboration & approval process. Design By, Engineering Approval, Manufacturing Approval, & Quality Approval.

GD&T symbols can be customized with feature modifiers like MMC (Maximum Material Condition), LMC (Least Material Condition), and Projected Tolerance Zone (P). These modifiers offer flexibility, but when misused or unsupported, they introduce more confusion than clarity.

• Using MMC on non-size features: MMC and LMC only apply when a feature has size. Applying these to planes or center planes without an actual size violates GD&T rules. Little trick I learned is if you can measure the feature with the OD jaws or ID jaws of a caliper, it’s a feature of size. Not the depth or step part of the caliper though!
• Missing boundaries for orientation or profile tolerances: Orientation and profile controls must have appropriate datums and modifiers that define boundaries. Without them, interpretations vary between machinists and inspectors.
• Using “M” or “L” without functional justification: Just because MMC gives bonus tolerance doesn’t mean it should always be used. If the bonus tolerance leads to a part being functionally unacceptable, it defeats the purpose.

Fix: Only use modifiers where they make functional sense. If bonus tolerance is not needed or could lead to misfits, leave it out. Double-check that every modifier used is supported by the feature’s geometry and role in the assembly.

GD&T is powerful, but complexity doesn’t always equate to clarity. I’ve seen the best prints covered in GD&T and it makes sense, is applied properly, and measurements are attainable on a CMM. This doesn’t make the print “over engineered.” I’ve seen the worst prints littered with misused GD&T or no GD&T at all. Many prints suffer from:

• Redundant controls: For example, calling out both flatness and parallelism, with the same tolerance, to the same surface adds no real value. This redundancy clutters the print and raises questions about inspection priorities. Now double stacked Feature Control Frames with a larger tolerance on parallelism so the tilt is not as controlled but Flatness having a tighter tolerance makes sense.
• Stacked or nested tolerance zones with unclear order of precedence: Composite position tolerances and profiles must clearly indicate whether datums apply simultaneously or separately. Misunderstanding the stacking order leads to entirely different interpretations.
• Over-tolerancing: Using unnecessarily tight tolerances increases part rejection and cost without improving performance. GD&T should optimize, not over-restrict. Let’s say I want to control the height of a part, tilt of the surface and the flatness. Well, I can put a .001 Profile call out and it’ll control it all. But if height can vary more than tilt, and tilt can vary more than flatness you can triple stack some Feature Control Frames. Profile of .010, Parallelism of .005 and Flatness of .001. This actually makes the part easier to make and inspect than a .001 Profile tight tolerance.

Fix: Use the minimum amount of GD&T necessary to communicate function. Consolidate controls where possible. When using composite or multiple segment frames, clearly distinguish between simultaneous and separate datum references.

Profile is one of the most versatile controls in GD&T, but it’s also one of the most misunderstood.

• No datums when they’re needed: A profile tolerance without any datum reference is treated as form-only. This may not reflect the designer’s intent if the feature should be located relative to the rest of the part.
• Wrong datums: Using a skewed datum reference frame can cause profile zones to misalign with the part’s functional surfaces.
• Incorrect tolerance zone visualization: Many engineers don’t fully understand how the profile tolerance zone applies in 3D, especially with complex geometries. This leads to incorrect setups on CMMs or manual gages.

Fix: Make sure to reference datums when the profile is used for location, not just form. Leverage CAD models or color mapping tools to visualize the tolerance zones during design and inspection.

Position is one of the most powerful GD&T tools, but also one of the most abused. Now it is a greater way to control hole locations instead of plus minus distances, but the design engineer needs to know they are controlling more than just a center of a diameter location.

• Treating it like linear dimensioning: Position must control a cylindrical tolerance zone in most cases, not a box-shaped window. Designers who don’t understand this often assume “±0.005” means the same thing as a Ø0.010 position tolerance. It doesn’t.
• Failing to apply the correct datum precedence: Many features are located with respect to datums that don’t constrain the part adequately in all six degrees of freedom, leading to gaging or CMM setup errors.
• Incorrect tolerance stacking: Position is often used without considering cumulative effects of datum feature tolerances and part size variation.

Fix: Ensure everyone involved—designers, machinists, and inspectors—understands what the position symbol is controlling and how to properly measure it. Use basic dimensions, not toleranced dimensions, when applying position controls.

Basic dimensions are theoretically exact and critical to understanding the tolerance zone location for most GD&T symbols. But many prints include errors such as:

• Toleranced dimensions instead of basic: This is a telltale sign that the designer may not fully understand GD&T. A toleranced dimension alongside a feature control frame sends mixed signals.
• Incomplete or missing basic dimensions: Without complete basic dimensions from all referenced datums, there’s no way to evaluate whether a feature is within tolerance.
• Unclear origin of basic dimensions: Especially when coordinate systems shift across views, it’s not always obvious where the basic dimension is measured from.

Fix: Treat basic dimensions as sacred—they define the location of the tolerance zone and must be fully clear and unambiguous. Use centerlines, datums, and clear origin indicators to support them.

Perhaps the biggest problem isn’t the mistake itself, it’s the lack of communication when those mistakes arise. Quality inspectors and CMM programmers often see the same issues repeatedly but don’t have a formal way to report them to engineering.

• Designers don’t get feedback: Inspection teams catch issues or interpret ambiguities differently, but that information rarely makes it back to the CAD model.
• Lack of training across roles: Many programmers and machinists are forced to “fill in the blanks” because they haven’t been formally trained in GD&T, even though they’re tasked with interpreting it daily.
• Tribal knowledge of GD&T passed down from team member to team member can be a positive thing but in a lot of cases, critical details don’t get passed down and the game of telephone begins.

Fix: Encourage cross-departmental training and regular design-for-inspection reviews. Build a loop where inspection feedback can improve drawing standards over time.

GD&T is a language—and like any language, it only works when all parties understand it fluently. Missteps in datum selection, modifier use, and tolerance application can create bottlenecks that ripple from engineering to the shop floor and beyond. But these issues are avoidable with the right mix of functional thinking, collaborative communication, and ongoing education.

Whether you’re designing parts, programming a CMM, or performing hands-on inspection, spotting the red flags early can save time, money, and frustration. When done right, GD&T becomes more than a standard—it becomes a tool for uniting teams, improving parts, and delivering precision with confidence.