In the treatment of acute myocardial infarction, delayed reperfusion therapy is associated with increased mortality. Predictors of delay have been described for fibrinolysis but not for primary percutaneous transluminal coronary angioplasty (pPTCA). Therefore, we studied 40,017 consecutive patients with acute myocardial infarction who underwent pPTCA in the National Registry of Myocardial Infarction between June 1994 and April 2000. Median door-to-balloon times were calculated, and factors independently associated with a delay of >120 minutes were determined by logistic regression. The median door-to-balloon time among all patients was 111 minutes (interquartile range 84 to 152). The proportion of patients with a delay of >2 hours was greater among those aged > or = 65 years (49% vs 41%), women (50% vs 42%), patients with contraindications to fibrinolysis (60% vs 41%), and those without chest pain on admission (61% vs 43%, all p <0.0001). Delay was also more common with transfer from another hospital (87% vs 38%), with presentation outside the hours of 8 A.M. to 4 P.M. (51% vs 38%), and in hospitals performing <49 pPTCAs/year (47% vs 41%, all p <0.0001). The strongest independent predictor of delay was hospital transfer, along with non-daytime presentation and low-volume centers. Older age, female sex, and non-white race were weaker predictors. Both patient and hospital factors are associated with delay in pPTCA after presentation. These findings may help design treatment algorithms to minimize delay, thus improving the survival benefit of pPTCA. These results may also help design trials of combination reperfusion strategies.