Diabetic retinopathy is a common cause of blindness, and screening can identify the disease at an earlier, more treatable stage. However, rural individuals with diabetes may have limited access to needed eye care. The objective of this project was to demonstrate the feasibility of a diabetic retinopathy screening program using a state-of-the-art nonmydriatic digital fundus imaging system. The study involved a series of patients screened in primary care and public health locations throughout seven predominantly rural counties in eastern North Carolina. Images of each fundus were obtained and sent to a retinal specialist. The retinal specialist reviewed each image, recorded image quality, diagnosed eye disease and made recommendations for subsequent care. Of 193 volunteers with a history of diabetes mellitus, 96.3 percent reported that they were very comfortable or comfortable with the camera. Eighty-five percent of images were rated as good or fair by the retinal specialist. The retinal specialist also reported being very certain or certain of the diagnosis in 84 percent of cases. Image quality correlated highly with the certainty of diagnosis (Spearman's rank order correlation coefficient = 0.79; P < 0.001). The average time since the previous examination by an eye care specialist for diabetic subjects was two years. Approximately 62 percent of diabetic patients had diagnosable eye conditions, the most common of which was diabetic retinopathy (40.9 percent). In this convenience sample, African Americans, despite similar age and disease duration, were more likely to have retinopathy. Digital imaging is a feasible screening modality in rural areas, may improve access to eye care, and may improve compliance with care guidelines for individuals with diabetes mellitus.