When gripping an object the fingers assume a posture that depends upon the dimensions and shape of the object being held (Jones and Lederman, 2006). As the fingers are extended to encircle larger or less curved objects their palmar skin is stretched producing an enlargement of the receptive fields of local touch units, an effect known to reduce locognosic acuity (Cody et al. 2010). However, finger extension concurrently reduces creasing of the palmar skin at the interphalangeal joints, thereby increasing contact with the object, which is likely to cause a counteracting improvement in tactile acuity. To test whether differences in finger posture exert a net effect upon tactile length discrimination during a static (fingers are not moved over the object) gripping task, we have compared performance whilst holding objects of two differing curvatures. Twenty young adults (12 female, 8 male; aged 18-24 years) were studied. The ability of blind-folded participants to estimate the lengths of raised strips (1.5 mm height, 24mm width, 7 lengths of range 24-36mm at 2 mm intervals) on the surfaces of two differently-shaped objects (flat-sided box i.e. zero curvature and curved beaker of circumference 18.1cm) of similar weight, was quantified. The raised strips were positioned beneath the participant’s index finger as the object was gently gripped between the fingers and thumb of the non-dominant hand. In each trial, a reference length (30mm) strip was first presented promptly followed by a test strip of one of the 7 lengths. The participant had to state whether the test strip was `longer ‘or `shorter’ than the reference. Each test length was presented 10 times, in a randomised order, and the associated probability of the judgement `longer’ was calculated. The interval of uncertainty (IU, a measure of length discriminatory threshold) was determined from standard psychophysical functions (probability of judgement `longer’ versus test length) for the two gripped objects. Mean (SD) IU values for the flat (box) and curved (beaker) surfaces were, respectively, 4.76 (1.17) mm and 5.78 (1.62) mm. Statistical analysis (paired t-test) of IU indicated that IU was significantly (p<0.015) greater (length discrimination was less precise) for the curved than flat surface.We interpret these findings as demonstrating that the ability accurately to judge the dimensions of a held object depends on the shape of the object’s surface. In addition, the observation that tactile length discrimination was poorer when flexing a finger around a curved object than extending it over a flat surface suggests that a reduction in accuracy due to loss of skin contact area (as directly quantified using a modified fingerprinting technique applied in 4 of our participants) with creasing at the interphalangeal joints was a dominant causal mechanism.