The benefit of thrombolysis is dependent on time to treatment, but there is lack of evidence of this relation in patients undergoing primary percutaneous transluminal coronary angioplasty (PTCA). The hypothesis that the relation of time to treatment to mortality is dependent on patient risk was tested in a series of 1,336 patients who underwent successful primary PTCA and were stratified into "low-risk" and "not low-risk" patient groups according to the Thrombolysis In Myocardial Infarction criteria. After stratification, 942 patients (71%) were at not low risk, and 394 (29%) were at low risk. The 6-month mortality rate was 9.3% for not low-risk patients and 1.3% for low-risk patients (p <0.001). Among not low-risk patients, longer time to treatment was associated with increased age and a greater incidence of cardiogenic shock. Unadjusted mortality of the not low-risk patients increased from 4.8% to 12.9%, with increasing time to reperfusion up to 6 hours, whereas mortality of the low-risk group was constant, with an increased time to reperfusion. For the not low-risk group, the univariate analysis revealed a relation between time to treatment and mortality (odds ratio 1.35; 95% confidence interval 1.06 to 1.73, p = 0.017). Time to reperfusion was not an independent predictor of mortality at multivariate analysis. Mortality for not low-risk patients who undergo successful primary PTCA is related to the delay from symptom onset to treatment. The effects of other variables associated with a longer time to reperfusion may have a stronger impact on mortality, obscuring the incremental value of time to reperfusion at multivariate analysis.