In testing the effect of grading versus nongrading of formative evaluations on cognitive learning and affective behaviors of graduate nursing students the following hypotheses were formulated: The experimental (E) group will learn and demonstrate affective behaviors significantly more than control I (CI) and II (CII) groups. Subjects were 95 graduate nursing students. E group subjects (N = 38) were taught by means of FE. with grades. CI (N = 32) and CII (N = 25) groups were taught by means of lecture-discussion and FE without grades, respectively. The content covered conditions of learning and instruction in nursing. Pre- and posttests provided data on cognitive learning. Affective measures and demographic data were obtained only at the posttest. The hypotheses were tested by means of t test. Fisher's Exact Probability Test and Pearson were done to test distribution and relationships between variables. Results showed that: The E group learned significantly more than the CI group, but the CII group learned significantly more than the E group; all groups learned significantly; there were no significant differences among groups in affective behaviors although CII subjects scored highest. Implications were made for education of patients, students, and staff.