This study sought to compare procedural and clinical outcomes of percutaneous transluminal coronary angioplasty (PTCA) performed with 6F guiding catheters introduced through the radial, brachial or femoral arteries. Transradial PTCA has been demonstrated to be an effective and safe alternative to transfemoral PTCA; however, no randomized data are currently available. A randomized comparison between transradial, transbrachial and transfemoral PTCA with 6F guiding catheters was performed in 900 patients. Primary end points were entry site and angioplasty related. Secondary end points were quantitative coronary analysis after PTCA, procedural and fluoroscopy times, consumption of angioplasty equipment and length of hospital stay. Successful coronary cannulation was achieved in 279 (93.0%), 287 (95.7%) and 299 (99.7%) patients randomized to undergo PTCA by the radial, brachial and femoral approaches, respectively. PTCA success was achieved in 91.7%, 90.7% and 90.7% (p = NS) of patients, with 88.0%, 87.7% and 90.0% event free at 1-month follow-up, respectively (p = NS). Major entry site complications were encountered in seven patients (2.3%) in the transbrachial group, six (2.0%) in the transfemoral group and none in the transradial group (p = 0.035). Transradial PTCA led to asymptomatic loss of radial pulsations in nine patients (3%). Procedural and fluoroscopy times were similar, as were consumption of guiding and balloon catheters and length of hospital stay ([mean +/- SD] 1.5 +/- 2.5, 1.8 +/- 3.8 and 1.8 +/- 4.2 days, respectively). With experience, procedural and clinical outcomes of PTCA were similar for the three subgroups, but access failure is more common during transradial PTCA. Major access site complications were more frequently encountered after transbrachial and transfemoral PTCA.