For grasping, Ganel, Chajut, and Algom (2008) demonstrated that the variability of the maximum grip aperture (MGA) does not increase with the size of the target object. This seems to violate Weber's law, a fundamental law of psychophysics. They concluded that the visual representations guiding grasping are distinct from representations used for perceptual judgments. Weber's law is however only relevant for one component of the measurable variability of MGA, namely the variability in the sensory system. We argue that when looking at the relationship between object size and grasping, the gain (often called slope) governing the relationship between target size and MGA can be used as an approximation to estimate the contribution of sensory noise to MGA variability. To test the idea that differences in gain modulate the relationship between target size and MGA variability, we examined grasping under a variety of conditions. We found that gain varied quite significantly across different tasks, but irrespective of gain Weber's law could not be found in any of the grasping tasks. Instead we repeatedly found an inverse relationship between variability and object size, i.e. variability decreased for bigger objects. This trend may reflect the reduced biomechanical freedom found for movements at the end an effector's effective range of motion. MGA variability may thus be dominated by non-sensory factors and therefore may constitute a poor choice to estimate the variability of the visual signals used by the brain to guide our grasping actions.