The initial results of percutaneous transluminal coronary angioplasty (PTCA) in this institution were described previously and were comparable with national statistics despite the comparatively small number of cases performed by individual physicians. Two years later, the number and complexity of cases have increased significantly. In the present study, the experience of 3 physicians (group 1) who performed greater than 100 PTCAs (143 +/- 35, mean +/- standard deviation) was compared with that of 14 physicians (group 2) who performed less than 100 PTCAs (25 +/- 16) during a recent 12-month period. Group 1 and group 2 performed 430 and 351 PTCAs, with a 91 and 84% success rate, respectively (p less than 0.01). Group 1 and group 2 attempted dilatation of 595 and 444 narrowings with an 85 and 81% success rate, respectively (p less than 0.025). The differences reflect outcomes with "complex lesions," with which group 1 had a higher success rate than group 2 (81 vs 69%, p less than 0.05) and a lower complication rate (3.1 vs 7.5%, difference not significant). Outcomes with "simple lesions" were similar for the 2 groups (93 vs 90%). As a result, the conclusions of the previous study should be modified. The present data suggest that while low frequency operators can perform PTCA of "simple lesions" with quite satisfactory results in the setting of an institution in which large numbers of PTCAs are performed and in which an expert team is available for support, outcomes with "complex lesions" are likely to be better in more experienced hands.