Parasites frequently reduce the fecundity, growth, and survival of individual hosts. How often do these virulent effects reduce the density of host populations? Spectacular examples show that recently invaded parasites can severely impact host populations--but what about parasites persisting long-term in host populations? We have addressed this issue using a zooplankton host (Daphnia dentifera) that becomes infected with a fungal microparasite (Metschnikowia bicuspidata). We combined observations of epidemics in nine lakes over 6 years, fine-scale sampling of three epidemics, and a mesocosm experiment. Most epidemics remained small (<10% maximum prevalence) and exerted little influence on host densities. However, larger epidemics more severely depressed the populations of their hosts. These large/severe epidemics started and peaked earlier than smaller/benign ones. The larger epidemics also exerted particularly negative effects on host densities at certain lags, reflecting the delayed consequences of infection on fecundity reduction and host mortality. Notably, negative effects on the juvenile stage class manifested later than those on the adult stage class. The results of the experiment further emphasized depression of host density by the fungus, especially on the density of the juvenile stage class. Consequently, this common parasite reduces the density of host populations when conditions foster larger outbreaks characterized by an earlier start and earlier peak. Given these considerable effects on host density seen in a number of large epidemics, parasitism may sometimes rank highly among other factors (predation, resource availability) driving the population dynamics of these hosts.