A PCR was developed and applied for the detection of Clostridium botulinum type C in 18 avian and environmental samples collected during an outbreak of avian botulism, and the results were compared with those obtained by conventional methodologies based on the mouse bioassay. PCR and mouse bioassay results compared well (100%) after the enrichment of samples, but PCR results directly indicated the presence of this microorganism in six samples, while only one of these contained the type C botulinal neurotoxin before enrichment. The PCR assay was sensitive (limit of detection between 15 and 15 x 10(3) spores per PCR), specific (no amplification products were obtained with other clostridia), and rapid, since sonicated and heated samples provided enough template for amplification without any DNA purification. Eleven isolates of C. botulinum type C were recovered from mallards (Anas platyrhynchos), grey herons (Ardea cinerea), and mud during investigation of this outbreak.