Objective: To analyze behavioral adaptation of hemiplegic patients performing a complex manual task without time constraint. It was postulated that ipsilateral motor disturbance could not be observed after a hemispheric stroke. Design: Two manual tasks were used: (1) a new one, ''PigTail,'' required the patients to run a 3-cm-diameter copper ring in a wooden handle along a wavy copper wire without any time constraint; (2) the second task was the Nine-Hole Peg Test (NHPT). Setting: A hospital department of rehabilitation. Patients: A consecutive sample of 36 patients, who had all suffered an ischemic stroke in the middle cerebral artery territory, 18 with left hemisphere damage (LHD) and 18 with right (RHD), and who had similar ages (mean 54+/-13), stroke severity, time since stroke (mean 60 days), and functional independence according to the FIM. Main Outcome Measure: Main data were number of faults, time in seconds, and difference of time for two trials. Analysis compared the results with the same hand for patients and 86 healthy subjects. Results: Patients scores for NHPT were worse than controls, whatever the side of the lesion (p <.05). For Pig-Tail, the number of faults by patients was greater than by controls (p < .05); time was higher for LHD, but not significantly, and was similar to controls for RHD. All patients and controls speeded up between the two trials. Although the RHD were clumsy, they were always faster than LHD patients. Conclusion: There are ipsilateral motor disturbances in a complex manual task after hemispheric stoke, even without a speed constraint, and regardless of the hemisphere damaged. Further studies are needed to examine speed control that seemed impaired by right hemisphere damage and could explain clumsiness in these patients. (C) 1996 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation.