To assess the relationship between the volume of percutaneous transluminal coronary angioplasty (PTCA) procedures performed in a cardiac catheterization laboratory and major complications after adjusting for case mix and to evaluate the applicability of current guidelines for minimum laboratory volume. Cohort study using the 1992 and 1993 registries of the Society for Cardiac Angiography and Interventions. Forty-eight cardiac catheterization laboratories from throughout the United States and Canada. All 19 594 consecutive patients without an acute myocardial infarction (MI) undergoing a first coronary balloon angioplasty. Emergency bypass surgery, MI, or in-hospital death. There was a significant decrease in the rates of in-hospital mortality (P = .04), emergency bypass surgery (P < .001), MI (P - .001), and major complications (defined as one or more of these outcomes; P < .001) with increasing cardiac catheterization laboratory volume. After adjustment for case mix using multivariable analysis, these associations persisted, although the association with mortality was no longer statistically significant. There was no significant difference in outcomes in laboratories performing at least 200 vs fewer than 200 procedures per year, the currently recommended minimum laboratory volume (odds ratio [OR] for major complications, 0.81; 95% confidence interval [CI], 0.53 to 1.25). However, a statistically significant decrease in major complications was observed in laboratories performing more than 400 procedures per year (adjusted OR, 0.66; 95% CI, 0.46 to 0.96; P = .03; and OR, 0.54; 95% CI, 0.38 to 0.78; P = .001) when laboratories performing 400 through 599 procedures and at least 600 procedures per year, respectively, are compared with those performing fewer than 200 per year. An inverse association between cardiac catheterization laboratory procedure volume and major complications during PTCA exists independent of differences in patients' risk profiles. Our data suggest that the currently recommended minimum laboratory volume may be too low to distinguish higher-risk from lower-risk laboratories.