To assess the role of blockade of beta-receptor subpopulations in the treatment of neuroleptic-induced akathisia (NIA), the specific beta 2-antagonist ICI 118,551 was compared to placebo in a double-blind study. After a baseline evaluation on placebo, patients were treated with ICI 118,551 or placebo. Five of six patients treated with ICI 118,551 showed improvements in NIA, while only one of four patients improved on placebo. Patients were then treated openly with propranolol, a mixed beta 1, beta 2-antagonist. Compared to ICI 118,551, no further improvement on objective measures of akathisia was seen on propranolol. Mean subjective assessments of NIA declined on propranolol, but changes were variable and not statistically significant.