A pair wise-matched (1:1) case-control study was performed in IGU patients. The aim of the study was to estimate the excess length of stay attributable to nosocomial infection (Nl) and the extra costs of the excess length of stay by comparison with hospital stay of infected versus uninfected patients using the matching method. The population of the study were all those patients that were admitted and stayed >48 hours in the IGU during the study period. Gases were defined as patients with nosocomial infection; controls were selected according to matching variables. Matching variables were primary diagnosis for admission, age and length of stay before the day of onset of the first infection. In crude comparison of the length of stay between patients with and without Nl, the excess length of ICU stay attributable to Nl was 14 days and between patients surviving infection was 16 days. The extra charges for excess length of ICU stay were 3005.12 euro per survivor. In matched-pair comparisons the length of stay attributable to Nl was 8.5 days and between patients surviving infection was 13 days. The extra charges for excess length of ICU stay were 2441.66 euro per survivor. In conclusion, matched-pairs comparison may proved to be a more appropriate and reliable analysis method than crude comparison for the purpose of evaluating the excess length of stay attributable to Nl and the extra charges for excess length of stay in ICU.