Tears in the peripheral vascular zone of 71 menisci in 68 knees were repaired by us from 1978 to 1986. The meniscus repair was done by open arthrotomy in 26 cases and by arthroscopic techniques in 45 cases. We have assessed the relative efficacies of open and arthroscopic repair techniques. The results were compared in knees with and without anterior cruciate laxity. The indications for meniscal repair included unstable peripheral detachments and longitudinal tears of the outer third of the meniscus. Open repair was performed by a posteromedial arthrotomy incision. Arthroscopic repair was done using the double-lumen guide system with a limited posterior incision for retrieval of needles. We have found that the arthroscopic technique is easier to perform than the open repair because some tears are too far inside the rim to lend themselves to open suture. The average follow-up is 4 years, 2 months, with a range of 2-10 years. There have been no neurologic or vascular injuries from either technique. Twenty-five patients have had a repeat arthroscopy. The overall failure rate was 9.8%. The difference between the failure rate of 11% in the open-repair group and 8.8% in the arthroscopic repair group was not statistically significant. The failure rate in anterior cruciate-stable knees was 8% versus a 13% failure in cruciate-deficient knees. We conclude that both open and arthroscopic meniscus repair techniques are safe and effective with few complications in both stable and unstable knees. Anterior cruciate ligament stability is ideal, but it is not mandatory for a successful result.