How does the nervous system control individuated finger movements, like those people use in playing the piano or performing delicate surgery? Independent movements of different fingers commonly are assumed to be generated by different muscles, and controlled from different parts of the primary motor cortex, not unlike the separate fingers, motors and software channels used for a robotic hand. But in fact, real fingers do not move independently. Passive mechanical linkages account for only part of the interactions between the digits. Additional interaction results from the multitendoned muscles that flex and extend the fingers. These muscles act simultaneously on multiple fingers, requiring contraction of additional muscles to check unintended movements of other digits when a given digit is moved. In the primary motor cortex (M1) hand area, single neurons typically are active during movements of more than one finger, and neurons throughout the M1 hand area are active no matter which finger moves. Rather than groups of physiologically similar neurons being organized somatotopically to control particular muscles or fingers, control of any finger movement thus involves a physiologically diverse subpopulation of neurons distributed throughout the M1 hand area. The subpopulation controlling a particular finger movement shares members with, and spatially overlaps, the subpopulation controlling any other movement. These findings suggest that individuated finger movements are controlled by a spatially distributed network of highly diverse neuronal elements in the M1 hand area.