Waterfowl botulism is unique among intoxications because toxin produced within its victims leads to secondary poisoning of other birds. Because of this phenomenon, the epizootiology of the carcass-maggot cycle of botulism resembles that of an infectious disease and the reproductive rate (R) of the disease could be defined as the average number of secondary intoxications attributable to a single carcass introduced into a marsh. I propose that toxin production and botulism occur commonly at a low level in many marshes and that factors which influence R determine when the disease expands into a large epizootic. A model that incorporates the number of carcasses occurring in a marsh, the probability of a carcass containing spores, the probability of a carcass persisting until toxin-bearing maggots emerge, and the contact rate between live birds and toxin, may be useful for predicting the extent of secondary-poisoning, for identifying questions for research, and as a theoretical basis for management.