Capture–mark–recapture methods were used to study the influence of density-dependent and density-independent factors on population dynamics of the spiny pocket mouse Heteromys desmarestianus in a lowland forest in Belize. Fourteen trapping sessions were conducted in six 0.5-ha grids at irregular intervals from July 2000 to March 2005. Responses by H. desmarestianus to four disturbance events depended on the type and magnitude of the events. Although fluctuations in density were strongly related to fruit availability, the disturbances tended to magnify direct and delayed density-dependent effects on reproduction and juvenile survival. A catastrophic flood in July 2000 eliminated the entire population, but by September 2001 population density had increased from 34.4 individuals ha−1 immediately pre-flood to 42.5 individuals ha−1. Indirect effects from a hurricane in October 2001 had greater and longer lasting influences on demographic rates than direct effects from less severe floods in August 2002 and September 2003. Fruiting ceased for almost 2.5 y after the hurricane, resulting in extremely low juvenile survival and a decline in density from 46.8 individuals ha−1 in January 2002 to 23.0 individuals ha−1 in January 2004. Fruiting began again in January 2004, and after several subsequent pulses of reproduction and generally higher rates of juvenile survival the population reached a maximum of 77.3 individuals ha−1 in March 2005. Inferences from the study, especially on the duration of delayed-density-dependent effects, must be made cautiously because of the irregular trapping intervals resulting from the unplanned disturbances. However, the results indicate that population dynamics of many tropical small-mammal populations will be driven by direct and indirect effects from density-independent events that amplify density-dependent influences on demographic rates.