We reviewed collapse (sudden onset of pallor, limpness and hyporesponsiveness) following the first infant (DPTP+Hib) vaccination reported to the enhanced passive surveillance system of the Netherlands in 1994-2003. All 1303 reports identified by the current RIVM (National Institute for Public Health and Environment) case definition were captured by the Brighton Collaboration (BC) case definition, with in 17 (1.3%) reports insufficient information. Over the years the proportion of the highest level of diagnostic certainty (level 1) increased due to more complete data from 70% to over 90%. We checked the BC case definition also on a sample of cases (with pallor or hyporesponsiveness) not meeting RIVM's case definition for collapse at the time. Sixty out of 200 cases were captured by BC but again rejected by RIVM. The sensitivity BC levels 2 and 3 appeared too high. We recommend a more restrict case definition by the Brighton Collaboration with certain exclusion criteria to make it more specific. Furthermore a change in the specifications for levels 2 and 3 will increase specificity and accommodate for the loss of sensitivity.