Investigated the reactivity of self-monitoring (SM) of alcohol consumption in normal drinking undergraduates assigned to either an SM (n = 40) or Recall only (n  = 29) procedure. Ss completed a retrospective timeline (TL) recall report of their drinking on 2 occasions. A between-groups analysis of TL data found no significant effects, suggesting that SM was not reactive. For the SM period, significantly more drinking was reported by Ss using the SM method. Data collected by the 2 methods differed in terms of reported frequency but not quantity of drinking, suggesting that some Ss forgot some drinking occasions when recalling their drinking by TL. While both methods provide comparable estimates of amount of alcohol consumed per drinking occasion, SM provides a more valid measure when either frequency or total amount consumed is important.