In Brazil, rabies is an endemic disease with a fatality rate of 100%. The number of cases has decreased, but the number of cases for treatment after exposure and treatment dropout is still high. This study investigated the causes of anti-rabies treatment abandon, after exposure, in Porto Alegre (RS, Brazil), from July to December 2006. A case series was designed. Two hundred and eighty abandon cases were selected through randomized systematic sampling, out of 962 registered in Sinan. Data was collected in people's homes through interviews, by means of a questionnaire. According to the interviewees, 66.4% concluded the prescribed number of vaccines. This information was not registered in Sinan. Among the subjects confirmed of abandoning the treatment (94/280), 24.5% reported that they thought it was not necessary to complete the treatment, while 13.8% felt that they did not receive clear guidelines about what to do. Health services attempted to contact absents in only 19.2% of the cases. Data entered in Sinan present failures. These occurred because patients started treatment at one health service and continued in a different one. As a consequence, information about the conclusion of the treatment was not entered into the system.