The influence of stent diameter in a direct stenting technique was analyzed. We retrospectively identified 987 consecutive lesions in 773 patients in whom direct stenting was attempted. Lesions were divided into two groups: group 1, nominal stent diameter 2.5 mm (237 lesions) and group 2, > or =2.75 mm (n=750). Differences between groups were found in age (64.4 [10.4] vs 62.3 [11] P=.009), female sex (33.2% vs 17%; P<.0001), diabetes (44% vs 33.1%; P=.003), tortuosity (5.4% vs 2.5%; P=.034), reference diameter (2.5 [0.3] vs 3.3 [0.6]; P<.0001) and location in distal segments (44.5% vs 29.4%; P<.0001). Primary deployment (85.5% vs 95.5%; P<.0001) and postdilatation success rates (1.9% vs 4.8%; P=.039) were higher in group 2, with no differences in vessel dissection rate (4.7% vs 4.4%; P=.85). Direct stenting with 2.5 mm stents was associated with a lower success rate than larger stents. Vessel tortuosity, angulation, calcification, lesion severity and distal location were also associated with a higher failure rate. The predictive power of our model was 0.87 (95% CI, 0.82-0.92).