We compared the effects of different "mastery" criteria and doses on the acquisition and maintenance of sight words for 4 second graders with and without disabilities. First, we replicated Set Analysis and Operant Analysis conditions where participants were taught sight words in 20-trial (4 operants, 5 opportunities) sessions. Acquisition criteria were applied to a set of 4 operants or to individual operants, respectively. Second, we extended the literature by evaluating a lower dose of the Operant Analysis condition in a 12-trial session (4 operants, 3 opportunities). Thus, we compared 3 conditions--Set Analysis 5, Operant Analysis 5, and Operant Analysis 3. All participants acquired novel sight words fastest and required far fewer teaching trials to maintain each sight word under Operant Analysis 3 compared with Set Analysis 5 and Operant Analysis 5 conditions. Implications for arranging acquisition criteria and the interaction with trial-dosages are discussed.