The population dynamics of Heligmosomoides polygyrus were studied in outbred male MF1 mice subject either to primary or repeated experimental infection. Little variability in susceptibility was observed between mice, but heterogeneity increased with both duration and intensity of primary infection; this result indicates that there are differences in parasite survival between hosts. The rate of parasite-induced host mortality was 4 X 10(-4) per parasite per host per parasite lifespan. The mortality rates of male and female larvae during their development in the intestinal wall were estimated as 0.033 and 0.021 per parasite per day respectively, and estimates of the expected lifespans of the adult male and female parasites in primary infection of 11.22 and 9.92 weeks were obtained. Approximately 40% of female worms were observed in copula at any one time, although this proportion was significantly depressed in hosts harbouring fewer than 50 parasites and during the first four weeks of infection. Parasite fecundity was markedly age-dependent; each female worm produced approximately 31,000 eggs during its lifespan. No density dependence in either worm survival or fecundity in primary infection was apparent. The only detectable effect of worm density was in association with spatial distribution in the intestine; high levels of infection were associated with a posterior shift in the location of a proportion of the parasite population. Characterization of the dynamics of primary infection allowed predictions to be made about the expected dynamics of repeated infection. The comparison of predicted results and observed data revealed unequivocal epidemiological evidence for the density-dependent regulation of parasite population growth during repeated infection, affecting both parasite survival and parasite fecundity. The results also demonstrated the existence of two types of host individual in which the dynamics of repeated infection were markedly different. It is concluded that immunological differences between mice (possibly under genetic control) may be responsible for the observed effects; approximately 25% of MF1 mice seem unable to generate any protective immunity against H. polygyrus, whereas 75% become almost completely refractory to reinfection. This experimental system could be used for quantitative investigation of the impact of acquired immunity and genetic heterogeneity on helminth population dynamics. Both are of obvious relevance with respect to the control of infections of medical and veterinary significance.