The Red Queen hypothesis predicts that sexual reproduction should be favoured when the risk of infection by parasites is high. However, this prediction may also be achieved independently by means of the reproductive assurance and the density-dependent transmission hypotheses because higher densities increase the chances of fertilization (i.e. mates are easier to find) and increase the risk of infection. Additionally, the Red Queen hypothesis makes a long-term prediction that infection should be genotype-specific, with time-lagged dynamics. In the present study, we attempt to consolidate these two predictions by investigating whether sexuality and parasitism alternate in a time-lagged fashion on an ecological time-scale. In a 3-year study, we examined the time-lagged dynamics between male frequencies, infection levels (by Centrocestus sp.), and snail densities of four natural populations of sexual and asexual Melanoides tuberculata snails. The results obtained provide no evidence of either prediction of the Red Queen hypothesis. We found partial support for reproductive assurance in a single population, and could not discriminate between the three hypotheses in another population. The remaining two populations did not support any of the hypotheses in question. © 2008 The Linnean Society of London, Biological Journal of the Linnean Society, 2008, 93, 537–544.