Using theoretical and experimental approaches, we examined whether blood pressure at the wrist can be accurately measured by a volume-oscillometric method using a small pad-type square cuff placed above the radial artery (RA). Nuclear magnetic resonance imaging of the wrist allowed us to determine the geometry of two-dimensional (2-D) finite-element models. Finite-element method (FEM) analysis predicted that the pressure transmission ratio (calculated tissue pressure over externally applied pressure; square cuff to RA) was 98.8% for a cuff with a bladder sidelength 0.25 times the wrist diameter placed on the skin surface between the tendons of the brachioradialis muscle (Ta) and the flexor capri radialis muscle (Tb) and over the site (L) at which RA crosses the most protuberant spot on the volar aspect of the distal end of the radius. In addition, FEM analysis using a 3-D finite-element model (constructed by extending the 2-D finite-element model at site L in the longitudinal direction) showed that for all square cuffs with bladder sidelengths greater than or equal to 0.25 times the wrist diameter, the external pressure was transmitted almost completely to RA beneath the cuff center. Moreover, when the bladder sidelength was 0.44 times the wrist diameter, but not 0.29 times the wrist diameter, the mean blood pressure measured at site L in human subjects was similar to that measured at the upper arm. Taken together, the theoretical and the experimental results suggest that (i) blood pressure at the wrist can be measured accurately using a pad-type square cuff placed on the skin surface between Ta and Tb at site L, and (ii) the minimum bladder sidelength for accurate readings is somewhere between one-third and one-half of the wrist diameter.