214 patients were studied in a randomised trial to determine whether administraiton of intravenous atenolol within 12 hours of chest pain reduced eventual infarct size, as estimated by cumulative enzyme release and by ECG changes. 135 patients already had ECG evidence of infarction at entry; 72 received atenolol which significantly decreased subsequent enzyme release (atenolol and control means = 121 IU, SE +/- 10 and 177 IU, SE +/- 17; 2p < 0.005) and enhanced R-wave preservation (atenolol and control means = 46% +/- 3 and 36% +/- 3; 2p < 0.02). 79 patients had no evidence of infarction at entry; 44 did not receive atenolol and 27 of these subsequently developed infarction, whereas only 11 of 35 treated patients infarcted during their hospital stay (2p < 0.01). In hospital, fewer atenolol patients died (4 vs 9), had non-fatal cardiac arrests (2 vs 6), or required therapy for heart-failure (36 vs 47). Unlike many previous trials which had negative results, in this trial we gave the drug intravenously and promptly (median of 4 hours from onset of pain to injecton), thereby achieving early beta-blockade.