Objective: To gain insight into the severity of pertussis in hospitalised cases. Methods: In 1998, hospitalisation data were collected through paediatric surveillance. Results: From 115 hospitalisation admissions collected, 55% of the patients were younger than 3 months of age and not vaccinated; 12% were 3-5 months of age and of these, 50% were incompletely vaccinated; 33% were 6 months and older and of these, 61% were vaccinated. Fourty-six percent of the patients were diagnosed as having pertussis, confirmed by a positive culture or PCR, and 44% by positive serology. Three unvaccinated cases less than three months old died. Cyanosis, apnoea, administration of oxygen, artifical respiration and bradycardia were more frequently reported for unvaccinated cases compared to vaccinated cases and the hospitalisation time was longer (median 10 days vs. 4.5 days). Although complications were not often reported for vaccinated cases, the only case with encephalopathy was vaccinated and 17% of vaccinated cases had pneumonia. Conclusion: The lower number of reported cases in 1998 seems to reflect the lower pertussis incidence that was also observed in routine surveillance of notifications and cases with positive serology. Like in 1997 pertussis was most severe and complications were more frequently reported in young unvaccinated infants less than three months of age. However, also typical and severe pertussis cases occurred among vaccinated individuals. Active monthly paediatric surveillance is useful to verify trends in routine surveillance; interpretation of the routine sources is hampered by changes in the notification law and decentralisation of serology.