1. Two monkeys were trained to grasp an object between the thumb and index finger and lift it to a vertical distance of 12-25 mm. Up to 12 different conditions defined by different combinations of object weights (15, 65, and 115 g) and four surface textures (oiled metal, smooth metal, fine and coarse sandpaper) were used. The apparatus was equipped to measure grip (prehensile) force, vertical (load) force, and object displacement. 2. The monkeys appropriately scaled the grip force for the weight and the coefficient of friction of the object. However, during the dynamic phase of the task (grasping and lifting), the monkeys increased the prehensile force in multiple steps, suggesting that they relied on sensory feedback from the fingers to attain an adequate grip force to lift the object rather than programming the lift in advance. 3. Single-unit activity of 248 neurons was recorded in the hand area of the primary motor cortex while the monkeys performed the task. Of 208 neurons tested for cutaneous and proprioceptive receptive fields (RFs), 96 were sensitive to cutaneous stimulation of the glabrous skin of the hand, whereas 82 received proprioceptive input from wrist and finger muscles. The concentration of neurons with cutaneous input was significantly greater in the rostral bank of the central sulcus compared with cells with proprioceptive RFs, which were more concentrated in the convexity of the precentral gyrus. 4. From the global sample, 199 cells were tested with the three object weights, and 128 of these with at least two surface textures were used in combination with the object weights. The discharge of 58/199 (29%) cells was modulated with the object weight. Cells with cutaneous (20/84, 24%) and proprioceptive (23/71, 32%) RFs were about equally responsive to the object weight. 5. A greater number of motor cortical neurons were influenced by surface texture than by object weight. Of 128 cells tested with at least two surface textures, 67 (52%) showed a modulation of their activity as a function of texture. A significantly greater proportion of neurons with cutaneous RFs (40/63, 63%) showed differential activity as a function of object texture than cells receiving proprioceptive input (21/47, 45%). 6. Weight- and texture-related neurons were not distributed equally in the rostrocaudal dimension of the motor cortex. Only 8% of texture-related cells were located in the convexity of the precentral gyrus, whereas 30% of weight-related neurons were recorded from this rostral zone.(ABSTRACT TRUNCATED AT 400 WORDS)