We describe the relations between active maintenance of the hand at various positions in a two-dimensional space and the frequency of single cell discharge in motor cortex (n = 185) and area 5 (n = 128) of the rhesus monkey. The steady-state discharge rate of 124/185 (67%) motor cortical and 105/128 (82%) area 5 cells varied with the position in which the hand was held in space ("static spatial effect"). The higher prevalence of this effect in area 5 was statistically significant. In both structures, static effects were observed at similar frequencies for cells that possessed as well as for those that lacked passive driving from the limb. The results obtained by a quantitative analysis were similar for neurons of the two cortical areas studied. It was found that of the neurons with a static effect, the steady-state discharge rate of 78/124 (63%) motor cortical and 63/105 (60%) area 5 cells was a linear function of the position of the hand across the two-dimensional space, so that the neuronal "response surface" was adequately described by a plane (R2 ≥ 0.7, p < 0.05, F-test in analysis of variance). The preferred orientations of these response planes differed for different cells. These results indicate that individual cells in these areas do not relate uniquely a particular position of the hand in space. Instead, they seem to encode spatial gradients at certain orientations. A unique relation to position in space could be signalled by the whole population of these neurons, considered as an ensemble. This remains to be elucidated. Finally, the similarity of the quantitative relations observed in motor cortex and area 5 suggests that these structures may process spatial information in a similar way