Weight and length profiles are presented of 3505 Ascaris worms recovered after mass anthelmintic treatment of a human community in January 1984 and, after an 11 month period of reinfection, in November 1984. Male and female worms recovered after reinfection were significantly heavier and longer than those expelled after initial treatment (P < 0·0001). Multiple regression models were employed to examine variability in parasite size. A positive influence of host body weight on the weight of parasites recovered in January, but not in November, was observed. No negative relationship was recorded at either date between worm size and the total number of worms harboured, even after controlling for host weight. Striking host age-related variability was observed in the distribution profile of weights and lengths of individual worms recovered in January. Children harboured predominantly smaller worms before initial treatment, while adults expelled mainly heavy worms. In contrast, worms expelled by both children and adults after reinfection were heavier and more homogeneous in size, particularly within the relatively heavily infected group. These patterns argue against a primary role for density- dependent or acquired resistance mechanisms in determining the size of Ascaris in humans. However, the results are consistent with a hypothesis initially suggested by Jung (1954) that established worms may inhibit the development of newly acquired Ascaris larvae, perhaps regulating their own abundance.