The Nurse Stress Checklist was developed to measure stress in clinical nursing as a multidimensional construct. The instrument was formulated within a transactional model of stress. The items were derived from five domains thought to contribute to nurse stress in clinical settings and were organized into a questionnaire. Holmes Schedule of Recent Events was included as a validity measure. The instrument was tested on 104 staff nurses working in three institutions in an urban community. Exploratory factor analysis was applied to the 74 items presented to subjects in Likert-type format. Five factors were derived and subjected to psychometric evaluation. Internal consistency reliability for the five factors was good, ranging from 0.80 to 0.91. Means and measures of dispersion supported the potential of the five subscales to discriminate among respondents on the attributes being measured. Intercorrelations of the factors provided evidence of the distinctiveness of the five components of stress, although factor loadings showed some overlap between Personal Reactions and Work Concerns and Work Concerns and Work Completion Concerns. Validity of the factors also was supported by correlations with Holmes' Schedule of Recent Events. Content validity was supported by comparison of these results with findings of other investigators. Limitations of the results are discussed, and recommendations for future work on the instrument are offered.