Introduction Good quality information regarding traffic injuries is hard to find and difficult to produce. A traffic accidents epidemiological surveillance system was designed in Neuqun, Argentina. Its implementation and evaluation are described.

Material and methods Interviews were carried out with 20 key informants in order to investigate on sources and information course. Eight focal groups were constituted to establish the case definition and the variables. An instrument was designed to obtain the information related to each event and its victims, which was incorporated in two related databases. The attributes of the system were evaluated by seven focal groups and a qualified professional discussion group.

Results Three sentinel sites and one sentinel unit were organised. 33 registration centres were established in total. Data were obtained regarding when and what time, where, and how, for each accident, vehicles involved, environmental factors, and number of victims. The victims info includes: age, sex, place of residence, role in the event, use of safety equipment, alcohol drinking and consequences. During the first year 453 events and 739 victims were registered. Risk maps for each region were obtained. The system was considered effective; its attributes (acceptability, simplicity, delivery and use of information, opportunity, sensitivity, flexibility, representativeness and positive predictive value) were highly qualified.

Discussion and conclusions A system built cooperatively by its own users has a great acceptance. Coordination with an action programme is essential. The experience was a very important antecedent for the implementation of the current National Injury Surveillance System.