The effects on memory and psychomotor functions of oxazepam (15, 30 mg) were compared with those of its chlorinated derivative lorazepam (1, 2 mg) and placebo. Forty five volunteers took part in a double-blind, independent groups design. Subjects completed a battery of tests before and 1.5 and 3 h after drug administration. Lorazepam and oxazepam had similar dose-related effects on tests of attention, manual motor speed and recoding skills and all active treatments produced similar levels of subjective sedation. Both drugs caused anterograde impairments of long-term verbal memory and had no effects on short-term verbal span or recency. Neither drug produced retrograde amnesia for material learned before drug administration nor affected retrieval from semantic memory. The high (2 mg) dose of lorazepam positively facilitated recall of pre-drug information. The magnitude of anterograde amnesia produced by the two benzodiazepines was not linearly dose-related in that the two low doses had similar effects whereas the high dose of lorazepam produced impairments many times greater than oxazepam 30 mg. We conclude that drugs with similar half-lives may have similar effects on some cognitive and mood factors but be completely different in terms of amnesic effects. Differing potencies of the two drugs may be a more important factor determining amnesic side-effects.