The safety and efficacy of a 10- to 15-mg/h continuous infusion of intravenous diltiazem were evaluated in 47 patients with atrial fibrillation or flutter who first responded to 20 mg or 20 mg followed by one or more 25-mg bolus doses of open label intravenous diltiazem. Of the 47 patients, 44 responded to the bolus injection and were randomized under double-blind conditions to receive either a continuous infusion of intravenous diltiazem (10 to 15 mg/h) (23 patients) or placebo (21 patients) for up to 24 h. Seventeen (74%) of the 23 patients receiving diltiazem infusion and none of the 21 with placebo infusion maintained a therapeutic response for 24 h (p less than 0.001). Over 24 h, patients receiving diltiazem infusion lost response significantly more slowly than did those receiving placebo infusion (p less than 0.001). Nonresponders to the double-blind infusion were given an additional bolus injection of open label intravenous diltiazem and administered an open label 24-h intravenous diltiazem infusion. The overall proportion of patients maintaining a response to a 24-h infusion of intravenous diltiazem under double-blind or open label conditions combined was 83% (34 of 41). Efficacy of the 24-h infusion of intravenous diltiazem was similar in elderly versus young patients, those who did versus those who did not receive digoxin and those weighing less than 84 versus greater than or equal to 84 kg. However, intravenous diltiazem appeared to be more effective in atrial fibrillation than in atrial flutter. No significant untoward effects were noted.(ABSTRACT TRUNCATED AT 250 WORDS)