The utility of an objective evaluation of motor activity, actigraphy, was examined in the evaluation of handedness. Hand preference was assessed in a homogeneous group of 190 young volunteers using the Edinburgh Inventory (EI). The EI distribution obtained in the population studied was comparable with distributions cited in the literature. Simultaneous actigraphic recordings from both wrists were made in 58 of these subjects for 20 h, starting at 1000 h, using an epoch length of 4 s. Care was taken to include comparable numbers of right- and left-handers (based on EI score) in this subgroup. Two actigraphic parameters were defined. One of these, the Activity Index (AI[x]), is a measure of the difference in total motor activity between right and left wrist. The other, Movement Index (MIy[x]) is a measure of the difference in movement pattern. AI[x] showed a moderate but significant correlation with EI (r = 0.36, P < 0.005). The correlation between MIy[x] and EI was high (r = 0.65, P < 0.0001). Rebinning of the data into 60 s epochs decreased the degree of linear correlation between MIy[x] and EI. We conclude, in contrast to a previous study, that actigraphy can be used to discriminate between dominant and non-dominant hands; that the difference in movement pattern between right and left hand is larger than the difference in total motor activity; and that epoch lengths shorter than the conventional 60 s are more sensitive for this kind of discrimination.