In this study, we have evaluated the performance of the Revised Trauma Score (RTS) as a triage instrument in the prehospital setting in The Netherlands. To this end we analysed prehospital and clinical data on 398 injured patients in an urban-rural area in the east of the Netherlands. Our study included injured patients aged over 15 who were alive at the time the ambulance arrived. We found a comparatively low prevalence of major injuries in the prehospital setting, which varied with the definition used (for patients with an HTI-ISS > or = 18, it was 5.8 per cent, for HTI-ISS > or = 20 it was 3.7 per cent; for a modified HTI-ISS criterion it was 5.3 per cent and 2.7 per cent needed major emergency therapy). Estimates of sensitivity were also rather low and varied with the definition used (38 per cent for HTI-ISS > or = 18; 56 per cent for HTI-ISS > or = 20, 45 per cent for the modified HTI-ISS criterion and 76 per cent for major emergency treatment). The specificity and the predictive value of a lowered RTS, however, were 94 per cent and 26 per cent respectively for all definitions used. The conclusion of this study is that the performance of the RTS in this study population is poorer than expected from earlier studies. The low prevalence of major injuries in the prehospital setting in The Netherlands and the distribution of case severity may possibly explain these results.