The effect and choice of a drug to control heart rate for symptomatic improvement in patients with isolated mitral stenosis with normal sinus rhythm (n = 10) or atrial fibrillation (n = 10) were studied. Digoxin (0.25-0.5 mg daily), metoprolol (50-100 mg twice a day) and verapamil (40-80 mg three times a day) were evaluated for this purpose. An open randomised cross-over design was followed. The efficacy of a drug was evaluated by: (1) subjective improvement on a visual analog scale, and (2) objective improvement on repeated multi-stage symptom-limited treadmill exercise. In patients with sinus rhythm greater than or equal to 50% subjective improvement was seen in 90%, 40% and nil with metoprolol, verapamil and digoxin, respectively. The total work done by these patients was 1008 +/- 541 kpm (control), 2869 +/- 1418 kpm on metoprolol, 2369 +/- 884 kpm on verapamil and 1654 +/- 918 kpm on digoxin. In patients with atrial fibrillation greater than or equal to 50% subjective improvement was seen in 80%, 40% and 30% with verapamil, metoprolol and digoxin, respectively. The total work done by these patients was 555 +/- 232 kpm (control), 1379 +/- 553 kpm on verapamil, 1251 +/- 575 kpm on metoprolol and 716 +/- 340 kpm on digoxin. The degree of improvement on a drug appeared to be a function of its ability to control resting and exercise heart rates in two different rhythms in these patients.(ABSTRACT TRUNCATED AT 250 WORDS)