The role of rapid health assessment in generating data other than routine reporting for a multi-element primary health care information system is presented. Rapid surveys, based on the adaptation of the WHO/EPI cluster survey methodology, may generate reliable and valid results useful for the support of a managerial PHC information system. However, because of the limitations inherent to the method, so far, only few studies have investigated more than few PHC related issues. The experience of a household rapid survey conducted in Arua District, Uganda, using a modified EPI cluster survey methodology, is reported. Rapid appraisal methods were used to prioritize the information requirement and to identify the survey items. Fully supervised teams of primary school teachers were used as interviewers. Data processing, check and analysis were speeded up by a lap-top computer, in spite of problems of erratic power. Within a 10-day time span between the start of the survey and the publishing of results, data on health services' utilization, health seeking behaviour, coverage of PHC services, including immunization, and anthropometric data on the nutritional status of under-five children were obtained. Standard errors and 95% confidence intervals were calculated taking into account the variability of the parameters under investigation and true design effects were computed. The findings were utilized for the identification of health priorities and the monitoring of effectiveness of programmes, as well as to validate routine reporting. The methodological package was built up looking at the local context, so that it could become an operational tool for the district health management team.