PURPOSE.: To evaluate knowledge, attitudes, and environment of primary care providers, and to develop a conceptual framework showing their impact on self-reported pre-school vision screening (PVS) behaviors. METHODS.: Eligible primary care providers were individuals who filed claims with Medicaid agencies in Alabama, South Carolina, or Illinois, for at least eight well child checks for children aged 3 or 4 years during 1 year. Responses were obtained on-line from providers who enrolled in the intervention arm of a randomized trial to improve PVS. We calculated a summary score per provider per facet: (1) for behavior and knowledge, each correct answer was assigned a value of +1; and (2) for attitudes and environment, responses indicating support for PVS were assigned a value of +1, and other responses were assigned -1. RESULTS.: Responses were available from 53 participants (43 of 49 enrolled pediatricians, 8 of 14 enrolled family physicians, one general physician, and one nurse practitioner). Recognizing that amblyopia often presents without outward signs was positively related to good PVS: [odds ratio (OR) = 3.9; p = 0.06]. Reporting that 'preschool VS interrupts patient flow' posed a significant barrier (OR = 0.2; p = 0.05). Providers with high summed scores on attitudes (OR = 6.0; p = 0.03), or knowledge and attitudes (OR = 11.4; p < 0.001) were significantly more likely to report good PVS behavior. There was a significant trend between the number of 'good' scores on knowledge, attitudes or environment, and 'good' PVS behavior (p = 0.04). CONCLUSIONS.: PVS is influenced by positive attitudes, especially when combined with knowledge about amblyopia. Interventions to improve PVS should target multiple facets, emphasizing (1) asymptomatic children are at risk for amblyopia, (2) specific evidence-based tests have high testability and sensitivity for amblyopia in pre-school children, and (3) new tests minimize interruptions to patient flow. Copyright © 2010 American Academy of Optometry.