We converted 158 Parkinson's disease (PD) patients on stable doses of standard carbidopa/levodopa (Std-L) to controlled-release carbidopa/levodopa (L-CR). Of the 141 patients who completed the study, 103 (73%) preferred L-CR, 26 (18.5%) preferred Std-L, and 12 (8.5%) had no preference. One hundred fourteen patients elected to continue L-CR, and we performed the primary data analysis on this group. Following conversion to L-CR, patients reported an increase in length of benefit from each dose and an increased "kick-in" time. There was a decrease in the total number of doses, "off" periods, sleep interruptions per night, dose failures, and sleep disturbances. Conversion to L-CR resulted in a significant increase in total levodopa dose. There was no significant change in the dyskinesias. However, early-morning dystonia resolved in eight of 14 patients. Our findings suggest that L-CR is particularly effective in decreasing motor fluctuations, reducing nocturnal problems, and minimizing levodopa dose failures in PD.