Although the validity of perceived exertion as a method of monitoring the intensity of resistance exercise has been established, little is known about the test-retest reliability of ratings of perceived exertion during resistance exercise. Specifically, it is unknown whether the use of different anchoring procedures influences the reliability of ratings of perceived exertion. 30 men were assigned to an Exercise, Memory, or combined Exercise and Memory anchoring group. Participants completed an assessment of maximal leg-extension strength and were introduced to the Borg 15-category rating of perceived exertion scale through anchoring procedures that varied across groups. During two sessions of resistance exercise, participants rated active muscle perceived exertion after performing one repetition of the leg-extension exercise at 40%, 50%, 60%, 70%, 80% and 90% of the one-repetition maximum. A three-factor (Group x Intensity x Session) analysis of variance was performed to examine the perceived exertion data. Perceived exertion increased significantly (p<.01) with increasing exercise intensity in all groups and in both sessions. Mean ratings did not differ significantly among groups. Reliability was assessed for each group. Intraclass correlation coefficients ranged from .07 to .80 and percent agreement ranged from 60% to 90%. The results indicate that the reliability of ratings of perceived exertion during resistance exercise is acceptable regardless of the type of anchoring procedures used.