Percutaneous transluminal angioplasty ± stent (PTA/S) and surgical bypass are both accepted treatments for claudication due to superficial femoral artery (SFA) occlusive disease. However, long-term results comparing these modalities for primary intervention in patients who have had no prior intervention have not been reported. We report our results with 3-year follow-up. We reviewed all lower extremity bypass procedures at Beth Israel Deaconess Medical Center from 2001 through 2009 and all PTA/S performed from 2005 through 2009 for claudication. We excluded all limb salvage procedures and included only those that were undergoing their first intervention for claudication due to SFA disease. We recorded patient demographics, comorbidities, perioperative medications, TASC classification, and runoff. Outcomes included complications, restenosis, symptom recurrence, reinterventions, major amputation, and mortality. We identified 113 bypass grafts and 105 PTA/S of femoral-popliteal lesions without prior interventions. Bypasses were above the knee in 62% (45% vein) and below the knee in 38% (100% vein). Mean age was 63 (bypass) versus 69 (PTA/S; P < .01). Mean length of stay (LOS) was 3.9 versus 1.2 days (P < .01). Bypass grafts were used less for TASC A (17% vs 40%; P < .01) and more for TASC C (36% vs 11%; P < .01) and TASC D (13% vs 3%; P < .01) lesions. There were no differences in perioperative (2% vs 0%; not significant [NS]) or 3-year mortality (9% vs 8%; NS). Wound infection was higher with bypass (16% vs 0%; P < .01). None involved grafts. Bypass showed improved freedom from restenosis (73% vs 42% at 3 years; hazard ratio [HR], 0.4; 95% confidence interval [CI], .23-.71), symptom recurrence (70% and 36% at 3 years; HR, 0.37; 95% CI, .2-.56), and freedom from symptoms at last follow-up (83% vs 49%; HR, 0.18; 95% CI, .08-.40). There was no difference in freedom from reintervention (77% vs 66% at 3 years; NS). Multivariable analysis of all patients showed that restenosis was predicted by PTA/S (HR, 2.5; 95% CI, 1.4-4.4) and TASC D (HR, 3.7; 95% CI, 3.5-9) lesions. Recurrence of symptoms was similarly predicted by PTA/S (HR, 3.0; 95% CI, 1.8-5) and TASC D lesions (HR, 3.1; 95% CI, 1.4-7). Statin use postoperatively was predictive of patency (HR, 0.6; 95% CI, .35-.97) and freedom from recurrent symptoms (HR, 0.6; 95% CI, .36-.93). Surgical bypass for the primary treatment of claudication showed improved freedom from restenosis and symptom relief despite treatment of more extensive disease, but was associated with increased LOS and wound infection. Statins improved freedom from restenosis and symptom recurrence overall.