Egg output in urinary schistosomiasis can be influenced by several factors, such as time of collection of urine, day to day variations, seasonal variations and environmental conditions. Additionally there are unavoidable random errors inherent in every test procedure. To assess the precision of Schistosoma haematobium egg counts, carried out after paper filtration, two measures are determined, namely repeatability (r) and reproducibility (R). 10 aliquots of 10 urine samples are examined by 5 microscopists and values for r and R are calculated following the methods of the British Standards Institution (BSI, 1987). Results show unexpectedly high values for the 2 measures in relation to the overall means in all 10 urine samples, and analyses of variance confirm that there are large variations between aliquots of equal volume drawn from the same urine specimen, highly significant in all 10 urine specimens; variations between microscopists are considerably lower and only highly significant in one urine specimen. The Poisson heterogeneity test for all urine samples refutes in 9 out of 10 cases the hypothesis that aliquots of equal volume taken from a well mixed urine sample could reasonably be drawn from a Poisson distribution. It is therefore concluded that egg counts are very valuable for community studies and should always be performed, but results must be interpreted with more caution than before and great care taken in analyzing counts (especially in longitudinal studies), in classifying them and in relating them to morbidity.(ABSTRACT TRUNCATED AT 250 WORDS)