The aim of this study is to report the arthroscopic subjective findings of reconstructed anterior cruciate ligament (ACL) with good clinical outcome. Graft used for reconstruction was either patella bone-tendon-bone (PBTB) autograft or hamstring tendon [quadrupled semitendinous and gracilis tendon (QSGT)] autograft. From March 1997 to September 2003, among 716 ACL reconstructions 209 knees (208 patients) were available for second-look arthroscopy at a mean 21.2-month (range, 14-70 months) postoperative period. The second-look arthroscopy focused on the evaluation of (1) continuity of the reconstructed ACL graft, (2) subjective graft tension using a probe, (3) the extent of synovial coverage, (4) the prevalence of cyclops or cyclops-like lesion, and (5) bony change after notchplasty. Patellar tendon autograft was used in 80 knees, hamstring tendon autograft in 129 knees. Just prior to second-look arthroscopy two objective clinical evaluations, KT-2000 arthrometer and Lysholm knee score, were performed to verify good clinical outcomes. A comparison between the hamstring tendon group and the patella tendon group, hamstring group showed slightly better results in Lysholm knee scores and KT-2000 arthrometer but there were no statistically significant differences (p>0.05). Undetected partial graft tear was seen in 21 knees (10%). With regard to graft tension, a total of 181 grafts (87%) showed normal tension and 28 (13%) showed slight lax tension. The overall synovial coverage was poor in nine (4%) knees. The synovial coverage was slightly better in the hamstring tendon group. A total of 45 knees (21.5%) showed cyclops-like lesion in variable sizes and locations. Reformation of the notch was seen in 85 knees (40%). In conclusion, the findings of second-look arthroscopy of reconstructed ACL in good clinical outcome patients showed approximately 10% partial graft tear, 5% poor synovial coverage, 20% cyclops-like lesion, and 40% some notch reformation.