The efficacy of nicorandil was compared with atenolol in 37 patients with chronic stable angina using a randomized, placebo-controlled, parallel study design. After a single-blind placebo phase, patients were randomized to receive nicorandil or atenolol using a double-dummy technique. Patients took nicorandil 10 mg twice daily or atenolol 50 mg once daily for the first 3 weeks, and if no adverse effects were encountered they took nicorandil 20 mg twice daily or atenolol 100 mg once daily, for the final 3-week phase. Treadmill exercise tests were performed at the end of each treatment phase immediately before and 2 hours after the morning dose of medication. Groups were demographically similar. Placebo exercise times were 7.06 (0.60) minutes (mean +/- standard error of the mean) in the nicorandil group and 6.81 (0.47) minutes in the atenolol group. After 6 weeks, improvements in exercise time were before dosing: +1.47 (0.40) minutes with nicorandil (p less than 0.005) and +1.33 (0.29) minutes with atenolol (p less than 0.001). Improvements after therapy was administered were +2.45 (0.41) minutes with nicorandil (p less than 0.001) and +2.37 (0.43) minutes with atenolol (p less than 0.0001). Whereas, the predose peak exercise double product (heart rate X systolic blood pressure mm Hg/100) was reduced with atenolol (-43.6 units; p less than 0.001), an increase (+7.56 units; difference not significant) was noted with nicorandil. One patient taking atenolol and 5 taking nicorandil developed persistent headaches. One subject with severe 3-vessel coronary artery disease had fatal myocardial infarction within 3 days of starting nicorandil, 10 mg twice daily.(ABSTRACT TRUNCATED AT 250 WORDS)