A 32-item Computer Self-Efficacy Scale (CSE) was developed to measure perceptions of capability regarding specific computer-related knowledge and skills. Bandura's theory of self-efficacy (1986) and Schunk's model of classroom learning (1985) guided the development of the CSE. Each of the skill-related items is preceded by the phrase "I feel confident." A five-point Likert-style response format was used. Data from 414 subjects who were learning to use computers in three settings were used to conduct analyses for assessing the reliability and construct validity of the instrument. The subjects included graduate students, adult vocational students, and nurses. Data were collected on beginning level self-efficacy, advanced level self-efficacy, mainframe self-efficacy, and willingness to exert effort. A principal factor analysis with oblique rotation produced a conceptually meaningful three-factor solution with high alpha reliabilities. Additional analyses provided some support for the theoretical propositions of self-efficacy, but suggested that the males and females in this study differed in judgments of their computer capability. Research using actual measures or observations of performance, effort expenditure, and persistence are needed to continue the exploration of the predictive validity of the three types of efficacy judgments generated by the CSE. Nine tables and one flowchart are included. (TJH)