Three examinations administered to medical students were analyzed to determine differences among severities of judges' assessments and among grading periods. The examinations included essay, clinical, and oral forms of the tests. Twelve judges graded the three essays for 32 examinees during a 4-day grading session, which was divided into eight half-day grading periods. Eighteen judges graded the performance of 217 examinees on the clinical examination during a 2-day grading session that was divided into four grading periods. Forty-six judges graded the performance of 270 examinees on the oral examination during a day and a half grading session that was divided into three grading periods. An extension of the Rasch model was used to analyze facets for examinees, items, judges, and grading periods. This study focused only on judge severities and differences among grading periods, however. The system of links necessary to calibrate judge severities and grading periods as separate facets was adequate because judges had 16 primary protocols and some examinees in common. Data from each of the three examinations were analyzed using FACETS, a computer program for Rasch analysis of examinations with more than two facets. The FACETS program estimates objective and conjointly additive calibrations, standard errors, and fit statistics for each element of each facet in the examination. Significant variation in judge severities and some variation across grading periods were found on all three examinations. However, the fit statistics confirm that most judges are reasonably consistent in the application of their individual level of severity. Four data tables and four graphs are included. (TJH)