Driving dimensions are the type of dimension that directly controls geometry. In CAD software, engineering drawings, and parametric modeling, a driving dimension sets a specific value, and the geometry conforms to it. Change the dimension, and the shape updates to match. This is distinct from other dimension types that simply report measurements without influencing the shape itself.
How Driving Dimensions Control Geometry
A driving dimension acts as the authority over a geometric feature. It dictates a specific value for length, angle, radius, or position, and everything in the model adjusts to obey it. If you set a driving dimension of 50 mm for a line in a CAD sketch, that line will be exactly 50 mm. Change it to 75 mm, and the line stretches automatically. The geometry is a direct output of the dimension’s value.
This is the core idea behind parametric modeling, which is the standard approach in tools like SolidWorks, Creo, Fusion 360, and other CAD platforms. Your entire 3D model is built on a network of driving dimensions. Editing any one of them ripples changes through the design, updating every feature that depends on it. This makes it easy to explore design variations or correct errors without redrawing anything from scratch.
Driving vs. Driven Dimensions
The key distinction is between dimensions that set rules and dimensions that report results. A driving dimension is the boss: it tells geometry where to go and how big to be. A driven dimension (sometimes called a reference dimension) is a follower. It measures what already exists but has no power to change it. Its value is calculated from the current state of the geometry, not the other way around.
For example, if you draw a rectangle with two driving dimensions for width and height, those two numbers fully define the shape. You could then add a driven dimension showing the diagonal length. That diagonal measurement updates automatically when you change the width or height, but you can’t edit the diagonal directly to reshape the rectangle. It’s purely informational.
In CAD software, driven dimensions are typically displayed differently, often in parentheses or a different color, so you can tell at a glance which dimensions you can edit and which are just for reference.
Dimensional vs. Geometric Constraints
Dimensions aren’t the only way to control geometry. CAD tools use two complementary systems: dimensional constraints and geometric constraints. Dimensional constraints handle measurable values like distance, length, angle, and radius. Geometric constraints handle relationships between objects, things like making two lines parallel, forcing a corner to be perpendicular, or locking a point to the center of a circle.
Autodesk recommends applying geometric constraints first to establish the overall shape of a design, then layering on dimensional constraints to pin down the exact sizes. Think of geometric constraints as defining the “type” of shape (a rectangle stays a rectangle, a tangent line stays tangent) while dimensional constraints define how large or small that shape is. Together, they fully lock down geometry so nothing is left ambiguous or free to move.
Degrees of Freedom and Full Constraint
Every geometric element starts with a certain number of degrees of freedom, meaning the number of independent ways it can move or change. A point on a 2D plane has two degrees of freedom (it can slide left/right and up/down). A rigid body in 2D has three: two for position and one for rotation. In 3D space, a rigid body has six degrees of freedom: three linear (movement along x, y, and z axes) and three rotational (roll, pitch, and yaw).
Each constraint you apply, whether dimensional or geometric, removes one or more of those freedoms. A dimension fixing a line’s length removes one freedom. A constraint pinning one end of that line to a specific point removes two more. When every degree of freedom has been eliminated, the geometry is “fully constrained,” meaning it has exactly one possible shape and position. Most CAD software will tell you when a sketch is fully constrained, under-constrained (still has some freedom to move), or over-constrained (has conflicting rules).
Basic Dimensions in Engineering Drawings
Outside of CAD modeling, there’s another important dimension type that controls geometry in a different way. In Geometric Dimensioning and Tolerancing (GD&T), the system used on manufacturing drawings to specify precision requirements, a basic dimension defines the theoretically perfect location or size of a feature. Basic dimensions appear enclosed in rectangular boxes on a drawing, and they never carry a tolerance value of their own.
A basic dimension represents your ideal CAD model, the exact point in space where a feature should be. It establishes where a tolerance zone is centered rather than defining the limits of acceptable variation directly. For instance, a basic dimension might place a hole exactly 25 mm from an edge. A separate geometric tolerance (like a position callout) then defines how far the actual hole can deviate from that perfect 25 mm target. The basic dimension controls the geometry of the ideal part; the tolerance controls how much reality is allowed to differ.
This two-layer system keeps drawings clean and precise. The basic dimension locks down the “true profile” or perfect target, and the tolerance zone wrapped around it handles the real-world variation that manufacturing inevitably introduces.
Practical Takeaway
If you’re working in CAD, driving dimensions are your primary tool for controlling geometry. They define sizes and positions, and editing them reshapes your model. Pair them with geometric constraints to fully lock down your design. If you’re reading an engineering drawing, look for basic dimensions (boxed numbers) as the features that define where geometry is supposed to be, with separate tolerances controlling how close it needs to get. In both cases, the principle is the same: the right dimensions don’t just describe geometry, they dictate it.