Theories of diagnosis in nursing (Carnevali, 1983; Gordon, 1982) and a model of intelligence (Guilford, 1979) were integrated to provide a theoretical basis for this study. In contrast to previous studies, accuracy of nursing diagnoses was measured as a continuous variable. It was hypothesized that three factors of divergent productive thinking, fluency, flexibility, and elaboration, would correlate positively with accuracy of nursing diagnoses. The sample consisted of 86 female nurses, graduates of generic baccalaureate programs with 1 to 5 years experience, who met the criterion of knowledge. Three written case studies (CS1, CS2, CS3) served as the criterion for accuracy. With CS1, none of the hypotheses were supported. With CS2, the three hypotheses were supported (p < 0.05) and 10% of the variance was explained by one factor, fluency (p < 0.01). With CS3, elaboration and accuracy were correlated (p < 0.05). The findings provide beginning support for inclusion of exercises for divergent thinking in nursing education and practice.