Recent studies showing a lack of internal consistency for the Holmes-Wright Type A Lantern (HWA) raise the issue as to whether the lantern test is repeatable. This study determines the HWA repeatability for several scoring criteria. There were 78 individuals with normal color vision (NCV) and 80 individuals with defective color vision (DCV) who were tested during two separate visits. Three runs of the test lights were always presented in dim room illumination. An additional dark-adapted run was presented during both visits for a subset of 61 subjects. Repeatability was evaluated for the following pass/fail criteria: the Joint Aviation Requirements, Farnsworth's criterion, the Commission Internationale de l'Eclairage criterion for the HWA, a perfect performance, and the 99th and 100th percentile scores of the NCV sample. The NCV agreement for passing both sessions was high, with 96% having a perfect score at both sessions. The DCV agreement for passing both sessions was lower, ranging from 0.62 to 0.80. Criteria which stopped the test after a perfect performance on the first run had the lowest values. The DCV agreement for failing both sessions was high, ranging from 0.97 to 0.99. Stopping the test after a perfect performance on the first run should be abandoned. Presenting three runs of the nine test lights and allowing no more than two errors is repeatable, passes all the NCV, and minimizes the number of individuals who pass the lantern at the first session and fail at a later session.