Fecal extracts and blood sera from 113 ducks showing clinical signs of botulism were examined for Clostridium botulinum type C toxin by means of the mouse toxicity test to evaluate coproexamination as a diagnostic procedure, as compared with demonstration of toxin in serum. When death of test mice unprotected with type specific antitoxin (while protected controls survived) was the criterion, 78.8% of the sera and 5.3% of the fecal extracts were positive. When characteristic signs of intoxication in the unprotected mice was included as evidence of toxin in the specimens, these percentages increased to 86.7 and 6.2, respectively. Fecal specimens were collected hourly for the first 6 h after peroral dosing of eight mallards (Anas platyrhynchos) with 1.0 LD50 of type C toxin and at 24, 48, and 72 h from birds surviving that long. From 2 to 4 toxin-positive specimens were passed by all eight ducks during the first 6 h, five specimens were positive at 24 h, and three were positive at 48 h. Only three specimens were collected at 72 h, all of which were negative. These findings suggest that attempts to detect toxin in the feces of wild ducks might have been more successful had the birds been captured earlier in the course of the disease.