OBJECTIVE: To provide epidemiological description of the cholera outbreak which occurred in Kampala between December 1997 and March 1998. DESIGN: A four-month cross-sectional survey. SETTING: Kampala city, Uganda. MAIN OUTCOME MEASURES: Number of cases reported per day, attack rate per age group and per parish, case fatality ratio. RESULTS: The cholera outbreak was due to Vibrio cholerae O1 El Tor, serotype Ogawa. Between December 1997 and March 1998, 6228 cases of cholera were reported, of which 1091 (17.5%) were children under five years of age. The overall attack rate was 0.62%, similar in the under-fives and five and above age groups. The case fatality ratio among hospitalised patients was 2.5%. The peak of the outbreak was observed three weeks after the report of the first case, and by the end of January 1998 (less than two months after the first case), 88.4% of the cases had already been reported. The occurrence of cases concentrated in the slums where the overcrowding and the environmental conditions resembled a refugee camp situation. CONCLUSION: The explosive development of the cholera outbreak in Kampala, followed by a rapid decrease of the number of cases reported is unusual in a large urban setting. It appeared that each of the affected slums developed a distinct outbreak in a non immune population, which did not spread to contiguous areas. Therefore, we believe that, a decentralised strategy, that would focus the interventions on each heavily affected area, should be considered in these circumstances.