This study compares long-term retention from audio-tutorial instruction with common lecture-discussion retention. The population of the experiment was limited to students in two sections of a teacher education methods course. Through the use of a univariate statistical analysis in the framework of a split-plot factorial design, no significant differences or interactions were found in either learning gain or retention. An indirect result of the study did indicate that there was no significant learning loss in the subject matter (tort liability) used in the experiment. While measurement error indications restricted the researchers from making bold inferences, it may be stated that valuable teaching materials evolved from the preparation of the tort liability lessons. (Author)