20 patients with Parkinson's disease and a fluctuating response to chronic treatment with conventional L-dopa preparations participated in an open randomized trial comparing two sustained-release L-dopa preparations (Madopar HBS, Sinemet CR4). While a majority (15 patients, 7 on Madopar HBS and 8 on Sinemet CR4) showed a favourable response after 2 months of slow-release L-dopa treatment the clinical benefit remained stable in only 2 patients on Madopar HBS and 3 patients on Sinemet CR4 over the entire follow-up period of 12 months. Reasons for treatment failure were increased peak-dose or biphasic dyskinesias or prolonged "off"-periods. This preliminary study failed to demonstrate clinically significant advantages of one slow-release principle over the other.