We studied the effect of diltiazem on mortality and reinfarction in 2466 patients with previous infarction from 38 hospitals in the United States and Canada. The patients were randomly assigned to receive diltiazem (240 mg per day, n = 1234) or placebo (n = 1232) and followed for 12 to 52 months (mean, 25). Total mortality rates were nearly identical among the two treatment groups (167 and 166, respectively), as were cumulative mortality rates. There were 11 percent fewer first recurrent cardiac events (death from cardiac causes or nonfatal reinfarction) in the diltiazem group than in the placebo group (202 vs. 226; Cox hazard ratio, 0.90; 95 percent confidence limits, 0.74 and 1.08). A significant (P = 0.0042) bidirectional interaction between diltiazem and pulmonary congestion was observed on x-ray examination. In 1909 patients without pulmonary congestion, diltiazem was associated with a reduced number of cardiac events (hazard ratio, 0.77; 95 percent confidence limits, 0.61 and 0.98); in 490 patients with pulmonary congestion, diltiazem was associated with an increased number of cardiac events (hazard ratio, 1.41; 95 percent confidence limits, 1.01 and 1.96). A similar pattern was observed with respect to the ejection fraction, which was dichotomized at 0.40. Thus, diltiazem exerted no overall effect on mortality or cardiac events in this population of patients with previous infarction. This neutral effect reflected a diltiazem-related reduction in cardiac events in the majority of patients without left ventricular dysfunction and an increase in such events in the minority of patients with left ventricular dysfunction.