1. This study was devoted to the neuronal processes underlying the construction of the motor program. Two monkeys were trained in a choice reaction time task to perform precise wrist flexion and extension movements of small and large extent. During a trial, the first visual signal, the preparatory signal (PS), informed the animal completely, partially, or not at all about direction and/or extent of the forthcoming movement. After a constant waiting period, a second visual signal, the response signal (RS), was illuminated calling for execution of the requested movement. 2. Reaction time (RT) and movement time (MT) measurements during the training as well as the recording sessions revealed that providing prior information about movement parameters strongly affected RT, but only slightly affected MT. Reaction time decreased in relation to the amount (number of movement parameters precued) and the type of prior information. Providing information about movement direction shortened RT much more than providing information about movement extent. Behavioral data support a parametric conception of motor programming, i.e., that the programming of the different movement parameters results from assembling separate processes of different duration. These results are compatible with the model in which programming processes are serially and hierachically ordered, movement direction being processed before movement extent. 3. Single-cell recording techniques were used to study neuronal activity of the primary motor (MI) and the premotor (PM) cortex, contralateral to the active arm. The activity of 155 neurons of MI and 158 neurons of PM was recorded during performance of the task. Of these 313 neurons, only 14 neurons did not change their activity during execution of the task. Two hundred and seven neurons whose activity changes were related to movement direction and/or movement extent have been selected for the further study. They were classified into three main groups: 1) execution-related neurons (49 in MI, 27 in PM), 2) preparation- and execution-related neurons (48 in MI, 54 in PM), and 3) preparation-related neurons (8 in MI, 21 in PM). 4. Directionally selective, execution-related neurons were found to be more frequently located within MI (81/105, 77.1%) than within PM (55/102, 53.9%), whereas directionally selective, preparation-related neurons appeared to be more frequently located within PM (47/102, 46.1%) than within MI (24/105, 22.9%).(ABSTRACT TRUNCATED AT 400 WORDS)