This work describes for the first time the incidence risk of passively reported canine rabies, and quantifies reported human exposure in N'Djaména (the capital of Chad). To diagnose rabies, we used a direct immunofluorescent-antibody test (IFAT). From January 2001 to March 2002, we were brought 34 rabies cases in dogs and three cases in cats. Canine cases were geographically clustered. The annual incidence risk of canine rabies was 1.4 (95% CI: 1.2, 1.7) per 1000 unvaccinated dogs. Most of the rabid dogs were owned-although free-roaming and not vaccinated against rabies. Most showed increased aggressiveness and attacked people without being provoked. Eighty-one persons were exposed to rabid dogs and four persons to rabid cats (mostly children<15 years old). Most of the exposed persons were neighbours or family members of the animal owner. Most exposures were transdermal bites, but nearly half of all exposed persons did not apply any first wound care or only applied a traditional treatment. In N'Djaména, humans are often exposed to canine rabies but do not use the full-course post-exposure treatment and wound care is insufficient. Most rabid dogs would be accessible to parenteral vaccination. Pilot vaccination campaigns are needed to determine the success of dog mass vaccination in N'Djaména as a way to prevent animal and human rabies.