This study investigated the relationship between engagement and achievement for college students in an intermediate swimming class. It also examined this relationship for students who entered the class with different initial skill levels, different previous experience with the subject matter, and for students of different gender. The methodology used is discussed under the following headings: subjects, instructors, skill rating, skill testing, instruction, and engagement coding. The original research premise is reported along with the results, which indicate that engaged time overall does not improve residualized achievement (i.e., students' breaststroke mechanics did not improve in correlation with the time they spent in class and practicing). Because the results did not agree with the hypothesis, four alternate hypotheses are discussed: (1) Engagement does not play a major role in predicting residualized gain; (2) The system for rating the skill was not sophisticated enough to measure relatively subtle aspects of improvement; (3) The method for coding engagement within the categories was not specific enough to detail important qualitative differences in the categories; and (4) Other factors mediate the power of engagement to predict residualized achievement. (JM)