Health promotion interventions intended to improve dietary behavior frequently incorporate self-efficacy as a construct to enhance behavior change. This paper presents results from a study to establish psychometric properties of a scale to measure children's self-efficacy for selecting healthful food. As part of a series of pilot studies to develop instrumentation for the Child and Adolescent Trial for Cardiovascular Health (CATCH), data were collected on third and fourth grade students (n = 1,127). Data analyses were conducted to estimate internal consistency, test-retest reliability, factorial validity, and criterion related validity. Results revealed acceptable estimates of internal consistency for the dietary self-efficacy scale (coefficient alpha = .84). Self-efficacy was strong associated with the children's usual food choices, accounting for about 34% of variance (Multiple R = .58). Findings support using such an instrument for evaluating intervention programs addressing nutrition behavior and for studies to determine the association of self-efficacy to dietary behavior or related constructs.