Twenty-nine patients with angina at rest took part in a randomized placebo-controlled short-term study to assess the relative effectiveness of different dosages of nifedipine (N), verapamil (V) and isosorbide dinitrate (ISDN) versus placebo and to evaluate the antianginal effects of a sustained-release preparation of ISDN (sr), of N retard form (r) and of V retard form (r). The 29 patients were divided into 3 groups: the first group of patients (10 patients, group A) was treated with N 10 mg six times daily, V 80 mg three times daily and ISDN 10 mg six times daily; the second group of patients (9 patients, group B) was treated with N 20 mg six times daily, V 120 mg four times daily and ISDN 20 mg six times daily; the third group of patients (10 patients, group C) was treated with N r 20 mg four times daily, V r 120 mg three times daily and ISDN sr 40 mg four times daily. The daily frequency of ischaemic episodes (IE) was assessed by Holter monitoring. The effect of each drug on the mean frequency of IE was compared with the placebo using a one-way analysis of variance and the Newman-Keuls test. In group A, the mean daily frequency of IE per patient was 8.1 +/- 5.9 with the placebo, 1.4 +/- 1.9 with N (P less than 0.001; -82%), 4 +/- 3.6 with V (P: NS; -50%) and 4.3 +/- 3.6 with ISDN (P: NS; -46%). In group B it was 6.4 +/- 3.4 with the placebo, 0.5 +/- 1.6 with N (P less than 0.01; -91%), 0.3 +/- 0.5 with V (P less than 0.01; -95%) and 1.2 +/- 1 with ISDN (P less than 0.01; -82%). In group C it was 10.3 +/- 8.7 with the placebo, 0.7 +/- 1.6 with N r (P less than 0.01; -93%), 1 +/- 2.5 with V r (P less than 0.01; -90%) and 5.1 +/- 7.7 with ISDN sr (P: NS; -50%). In group A a reduction of 100% in the number of recorded IEs was achieved in 5/10 patients by using N, in none by V, and in 1/10 by ISDN. In group B, in 8/9 patients by N, in 6/9 by V and in 3/9 by ISDN. In group C, in 8/10 patients by both N r and V r in 4/10 patients by ISDN sr.(ABSTRACT TRUNCATED AT 400 WORDS)