The central effects of two different formulations of Carbamazepine (CBZ) have been examined by analysis of saccadic eye movements (SEM) in 6 healthy volunteers, who took part in a double-blind, latin-square, placebo controlled study. Both a conventional- and a controlled-release formulation of CBZ produced a significant effect on peak saccadic velocity and saccade accuracy, but only the former affected saccade latency. Computer analysis of SEM confirmed it to be a highly sensitive method for detection of subclinical drug effects on the CNS.