A review of the literature on the influence of testing on learning and retention reveals the need for more comprehensive research findings in this regard. This study was designed to investigate further the direct effects of tests in contrast with no tests on learning and retention in ongoing college-level classes with both the instructor and the students going about the daily business of teaching and learning. It probed further to determine if it was taking the test or studying for it or both which effected learning and retention if indeed such effects were replicable. Test type (multiple-choice and short answer) and item complexity (knowledge and concept) were also variables studied. Moving from the cognitive domain to the affective domain, this study focused on the students' feelings when they did/did not have to study for and take a test. Analyses of the resulting data showed that students who studied for and took a test not only achieved more but also retained their learning longer than students who "studied in order to learn than for a test." However, student ratings of construction were lower when students were tested. (Author/AL)