The Lens Opacities Classification Systems (LOCS III) was developed and standardized using photograding. The purpose of this study was to assess the validity of LOCS III at the slitlamp and to compare slitlamp with photograding. To do so, two independent observers graded cataract at the slitlamp and in photographs from two sets of patients; the first set consisted of 205 eyes (193 acceptable photographs) and the second set of 51 eyes (51 photographs). The 95% tolerance limits (TL) for grading at the slitlamp ranged from 0.9 to 1.8 for the first set and 0.6 to 1.2 for the second (intraclass correlation coefficients 0.79 to 0.91 vs. 0.70 to 0.97, respectively). Specifically, there was a significant decrease in 95% TL for cortical and nuclear color. For the first set of photograding, the 95% TL were 0.3 to 0.6 between the two observers and 0.6 to 0.8 for the same observer at two different sessions. Similar results were found for photograding the second set. The 95% TL for comparing slitlamp and photograding were generally < 1.0. The results suggest that (1) LOCS III at the slitlamp has 95% TL only slightly worse than those for LOCS III photogradings; (2) LOCS III slitlamp grading for cortical and nuclear color improves with practice; and (3) the slitlamp and photographic gradings cannot be used interchangeably.