BACKGROUND: A trachoma control programme was started in southern Sudan in 2001. We did a 3-year evaluation to quantify uptake of SAFE (surgery, antibiotics, facial cleanliness, and environmental change) interventions, and to assess the prevalence of active trachoma and unclean faces. METHODS: Cross-sectional surveys, including clinical assessment of trachoma (WHO simplified system) and structured questionnaires, were done in four intervention areas at baseline and follow-up. Process indicators were uptake of SAFE components; primary outcome indicators included trachomatous inflammation-follicular (TF) and unclean face in children aged 1-9 years. FINDINGS: There was heterogeneous uptake of SAFE between intervention areas. Surgical coverage was low in all areas (range 0.5% of 428 individuals in Katigiri to 6% of 5002 in Kiech Kuon), antibiotic uptake ranged from 14% of 1257 individuals in Kiech Kuon to 75% of 954 in Katigiri, health education ranged from 49% of 190 households in Kiech Kuon to 90% of 182 in Padak, and latrine coverage from 3% of households in Tali to 16% in Katigiri. Substantial decreases in prevalence of TF and unclean faces were recorded in Katigiri and Tali, two of three sites where uptake of antibiotics and health education was high: TF decreased by 92% (95% CI 87-96) and 91% (86-95), respectively, and unclean face decreased by 87% (78-94) and 38% (22-52), respectively. Moderate effects were recorded in Padak, an area with high coverage, with a 28% (14-41) decrease in TF and a 16% (7-25) decrease in unclean face. No evidence of decline was seen in Kiech Kuon, where uptake of antibiotics and health education was low, with a 2% (-10 to 12) decrease in TF and a 10% (-3 to 23) decrease in unclean face. INTERPRETATION: Our results show that substantial falls in active trachoma can occur where SAFE is implemented, and that good results could be achieved with the SAFE strategy in other trachoma-endemic areas.