Periodic review and analysis of didactic achievement tests are essential parts of the process of improving teaching, learning, and testing programs. This study describes a method for determining the degree of cognitive congruence between teaching intentions, student expectations, and testing practices. A three-category model (recall, comprehension, and application) adopted from Bloom's Taxonomy was used to survey periodontology faculty and thirdyear dental students regarding their opinion of how test questions should be distributed. Actual distribution of test items used to evaluate student didactic achievement were also categorized by three raters. In addition, item difficulty and discrimination were calculated for each category of items andfor each course. Faculty recommended a distribution heavily weighted in favor of comprehension and applicationtype items, while the actual distribution and student recommendations favored recall-type items. Recall and comprehension questions were equal in terms of difficulty and discrimination. However, application items were less difficult and had the lowest discrimination. The approach used in this study resulted in information useful for planning future test construction.