Spain's Epidemiologic Notifiable Disease Surveillance System (ENDSS) was evaluated by reference to Brucellosis, the zoonoses with greatest incidence countrywide and a disease that is notifiable on a case by case basis in the Basque Autonomous Region (BAR). ENDSS information and the use to which it was put, were analysed. Active case searching was carried out via the Microbiologic Information System (MIS), public hospitals and the Occupational Disease Registry (ODR). A comparison was run between reported and search data. The capture-recapture method was used to compare ENDSS against hospital cases. Risks and outbreaks wee detected. Suspect cases were not reported. There was a 40% post-search increase in cases, i.e., 40 versus a previous figure of 24. Sensitivity was estimated at 60%. Predictive Value Positive could not be computed however. No case definition was in force. As estimated by the capture-recapture method, overall sensitivity for both ENDSS and hospital-based surveillance was 84% (95% IC: 70-100). The number of cases estimated by this method was 38 +/- 8. Underreporting is in evidence and, as a consequence, sensitivity is low. A case definition is called for. Response time is slowed by awaiting confirmation before reporting. The number of cases detected through active searching is compatible with that calculated by the capture-recapture method using only two sources.