Variations in fecal Schistosoma japonicum egg counts were studied in ZhuXi administrative village, JiangXi Province, China. Population stool examinations were collected with duplicate, standard, 41.5-mg Kato-Katz thick smears on seven consecutive days for 570 individuals from two natural (individual) villages: village I with high endemicity and village II with low endemicity. The proportion of individuals with at least one positive count increased from 42.4% after a single measurement to 68.3% after seven measurements in village I (n = 356), and from 17.0% to 36.0% in village II (n = 214), respectively. This demonstrates a very high variation in repeated S. japonicum egg counts and a considerable lack of sensitivity of the Kato-Katz technique; light and moderate infections are especially missed with a single or a few measurements. The observed day-to-day variation in individual egg counts is highly aggregated (variance higher than the mean) and suggestive of a negative binomial distribution. For five individuals on three days, repeated sampling from different locations of a stool specimen shows a clear trend with egg counts decreasing from the beginning of the stool to the end and from the outside layer to the center. Ten multiple samples from a particular subsection (10-30 g) of a stool specimen for 44 positive individuals still showed aggregation in egg counts, particularly for high intensities of infection. This means that the aggregation in repeated daily S. japonicum egg counts cannot be explained alone by a specific day-to-day component and variation in the concentration of eggs at different locations in the stool. There also exists clustering of eggs within parts of the stool.