A bilateral tolerance is a manufacturing specification that allows a part’s dimension to vary in both directions, larger and smaller, from a target size. It’s written as a nominal dimension followed by a plus-minus value, like 25.00 ± 0.05 mm, meaning the finished part can be anywhere from 24.95 mm to 25.05 mm and still be acceptable. This is the most common way tolerances appear on engineering drawings, and understanding how they work is essential for anyone reading blueprints or designing parts.
How Bilateral Tolerances Work
Every manufactured part has a target dimension, the size the designer ideally wants. But no manufacturing process is perfect, so engineers specify a tolerance zone: the range of sizes that will still function correctly. A bilateral tolerance splits that zone across both sides of the target.
The notation is straightforward. A dimension written as 36.95 ± 0.010 mm means the acceptable range runs from 36.940 mm to 36.960 mm. The “bilateral” part simply means the variation is permitted in both directions from the nominal value, not just one. Think of it as a target with equally sized margins of error above and below.
Equal vs. Unequal Bilateral Tolerances
Bilateral tolerances come in two forms. The most common is the equal (or symmetric) type, where the allowable variation is the same in both directions. A callout of 1.500 ± 0.001 inches permits the part to be 0.001 inches too large or 0.001 inches too small, with the target sitting exactly in the middle of the range.
The second form is the unequal (or asymmetric) bilateral tolerance, where the plus and minus values differ. For example, a 1 mm tolerance might be split as +0.80 / -0.20, meaning the part can grow 0.80 mm above nominal but only shrink 0.20 mm below it. The target dimension is no longer centered in the tolerance zone. Engineers use unequal bilateral tolerances when the design can handle more variation in one direction than the other, which is common with shaft-and-hole fits where a slight shift toward a tighter or looser fit is preferable.
On a drawing, unequal values are typically stacked vertically next to the dimension, with the plus value on top and the minus value on the bottom. Some CAD and drawing standards flag this with a “U” symbol to indicate an unequally disposed tolerance zone.
Bilateral vs. Unilateral Tolerance
The key alternative to a bilateral tolerance is a unilateral tolerance, where variation is only permitted in one direction from the target. A unilateral callout might read 1.499 +0.002 / -0.000, meaning the part can be up to 0.002 inches larger than 1.499 but cannot be any smaller. The entire tolerance zone sits on one side of the nominal dimension.
The choice between the two reflects design intent. Bilateral tolerances communicate that the target dimension is the ideal size and that deviation in either direction is equally acceptable. When parts are manufactured to an equally disposed bilateral tolerance, the expected distribution of finished parts centers around the target value. Unilateral tolerances, by contrast, deliberately shift that distribution toward one end of the range. This is useful when, for example, parts perform better at their maximum or minimum material condition. If an assembly fits better when a bore is slightly larger, a unilateral tolerance pointing in the plus direction tells the manufacturer exactly that.
How They Appear on Drawings
The simplest notation is the plus-minus format: the basic size followed by ± and a single value. This always indicates an equal bilateral tolerance. When the plus and minus values differ, they’re written separately, stacked above and below the dimension line.
An alternative approach, and the one preferred by the ASME Y14.5 standard, is limit dimensioning. Instead of showing a nominal value with tolerances, the drawing states the maximum and minimum sizes directly. A bilateral tolerance of 3.51 ± 0.02 would appear instead as 3.49 – 3.53 (or with the upper limit stacked above the lower limit). Both formats convey the same information. Limit dimensioning simply eliminates the math for the machinist, showing the acceptable range outright.
When Engineers Choose Bilateral Tolerances
Symmetric bilateral tolerances are the default for general-purpose dimensions where there’s no strong reason to prefer one direction of variation over the other. Lengths, widths, hole positions, and most features on a part will carry standard bilateral tolerances, often defined by a general tolerance block in the corner of the drawing rather than called out individually.
For precision fits between mating parts, like a shaft sliding into a bearing or a pin pressed into a hole, the choice gets more deliberate. Engineers frequently use unequal bilateral tolerances (or shift to unilateral tolerances) so the tolerance band aligns with the type of fit they need. A shaft intended for an interference fit, for instance, might carry tolerances that are both on the plus side of nominal, ensuring the shaft is always slightly oversized. Mathematically, you can always convert between symmetric and asymmetric notation by shifting the nominal dimension, but engineers often prefer to keep the nominal at the modeled CAD size and adjust the tolerance band instead, especially when a fit might change from clearance to interference during the design process.
Impact on Manufacturing Cost
The width of a bilateral tolerance zone directly affects how expensive a part is to produce. Tighter tolerances demand higher-precision machines and slower processes, which drives up cost per part. Looser tolerances allow faster, cheaper methods but increase the risk that assembled products won’t perform as well.
This creates a fundamental trade-off. Tight tolerances improve product performance and reliability. Loose tolerances reduce production cost. The cost relationship isn’t linear: as tolerances get very tight, the required process precision climbs steeply, and so does the price. Beyond direct machining costs, tight tolerances also increase scrap rates, because more finished parts fall outside the acceptable range and must be reworked or discarded. A well-chosen bilateral tolerance is wide enough to keep manufacturing practical but narrow enough to guarantee the part works in its assembly. Worst-case tolerance calculations, where every part is assumed to be at its extreme limit simultaneously, tend to produce tolerances tighter than necessary and inflate costs as a result.