This randomized trial compared four treatments for varicose great saphenous veins (GSVs). Five hundred consecutive patients (580 legs) with GSV reflux were randomized to endovenous laser ablation (980 and 1470 nm, bare fibre), radiofrequency ablation, ultrasound-guided foam sclerotherapy or surgical stripping using tumescent local anaesthesia with light sedation. Miniphlebectomies were also performed. The patients were examined with duplex imaging before surgery, and after 3 days, 1 month and 1 year. At 1 year, seven (5.8 per cent), six (4.8 per cent), 20 (16.3 per cent) and four (4.8 per cent) of the GSVs were patent and refluxing in the laser, radiofrequency, foam and stripping groups respectively (P < 0.001). One patient developed a pulmonary embolus after foam sclerotherapy and one a deep vein thrombosis after surgical stripping. No other major complications were recorded. The mean(s.d.) postintervention pain scores (scale 0-10) were 2.58(2.41), 1.21(1.72), 1.60(2.04) and 2.25(2.23) respectively (P < 0.001). The median (range) time to return to normal function was 2 (0-25), 1 (0-30), 1 (0-30) and 4 (0-30) days respectively (P < 0.001). The time off work, corrected for weekends, was 3.6 (0-46), 2.9 (0-14), 2.9 (0-33) and 4.3 (0-42) days respectively (P < 0.001). Disease-specific quality-of-life and Short Form 36 (SF-36(®)) scores had improved in all groups by 1-year follow-up. In the SF-36(®) domains bodily pain and physical functioning, the radiofrequency and foam groups performed better in the short term than the others. All treatments were efficacious. The technical failure rate was highest after foam sclerotherapy, but both radiofrequency ablation and foam were associated with a faster recovery and less postoperative pain than endovenous laser ablation and stripping.