OBJECTIVE To compare effects of immediate-release (IR) and sustained-release (CR) carbidopa/levodopa in levodopa-naive PD patients. BACKGROUND It was hypothesized that the long-acting preparation would be associated with fewer long-term complications. METHODS A total of 618 patients were studied in 36 centers worldwide in a blinded, randomized parallel study. Measures of efficacy and adverse reactions were recorded at 3-month intervals for 5 years. Motor fluctuations and dyskinesias were evaluated by a patient diary and a physician-recorded questionnaire. The Nottingham Health Profile (NHP) was used to evaluate quality of life. RESULTS Approximately 60% of patients completed the trial. After 5 years, the mean dose of IR was 426 mg/day, and the bioavailable dose of CR was 510 mg/day (mean dose, 736 mg/day). After 5 years, 20.6% of the IR group and 21.8% of the CR group had motor fluctuations or dyskinesia. Sixteen percent of both groups had changes in motor response by the questionnaire's definition. There was no significant difference between the two treatment groups. Disability scores and the motor score of the Unified Parkinson Disease Rating Scale (UPDRS) were highest at baseline, improved with therapy, and thereafter worsened over time to reach baseline scoring at the end of 5 years. The CR group was superior to IR for the Activities of Daily Living subsection of the UPDRS for all 5 years and for emotional reactivity and social isolation on the NHP; however, this may have resulted from higher doses of CR that were used. CONCLUSIONS Despite the progressive nature of PD, both the immediate-release and sustained-release carbidopa/levodopa formulations maintained a similar level of control in PD after 5 years compared with baseline. Additionally, the low incidence of motor fluctuations or dyskinesia was not significantly different between the treatment groups and may be partly attributed to the relatively low doses of levodopa used throughout the 5-year study.