A multicenter, double-blind, placebo-controlled, randomized trial of nifedipine, metoprolol and their combination was conducted in 338 patients with unstable angina (hospital admission diagnosis) who had not previously received treatment with a β blocker. In addition, nifedipine was compared with placebo in 177 patients who were receiving β blockers upon hospital admission. The main outcome event was the recurrence of ischemia or progression to myocardial infarction within 48 hours. Trial medication effects are expressed as ratios of event rates relative to placebo, e.g., for nifedipine as the event rate under nifedipine divided by that under placebo; 95% confidence intervals are also given. In patients not pretreated with a β blocker the rate ratio for nifedipine was 1.15 (0.83, 1.64), for metoprolol 0.76 (0.49, 1.16) and for the combination 0.80 (0.53, 1.19). In patients already receiving a β blocker, the addition of nifedipine was favorable and the rate ratio was 0.68 (0.47, 0.97). Equal numbers of patients developed myocardial infarction and reversible ischemia. Most infarctions occurred early, within 6 hours of randomization. In patients who were not already taking a β blocker, the nifedipine rate ratio for infarction only was 1.51 (0.87, 2.74). These results suggest that, in patients not previously receiving β blockers, metoprolol has a beneficial short-term effect on unstable angina, that a fixed combination with nifedipine provides no further gain and that nifedipine may be ineffective or counterproductive. On the other hand, the addition of nifedipine to existing β blockade when the patient becomes unstable seems beneficial.