The time course of serum concentration and performance on a concurrent probability matching task were evaluated in normal adults receiving 0.15 or 0.3 mg/kg of methylphenidate. The behavioral task, an arcade-like problem-solving game, revealed that drug-treated subjects improved their performance upon repeated testings during pharmacokinetic evaluation at a lower rate than did non-treated controls over the same time span. However, drug-treated subjects failed to adopt the adaptive problem-solving strategies selected by controls.