Between 1 September 2005 and 31 August 2006, a total of 148 sets of discrepant antimicrobial susceptibility data were reviewed on a monthly basis. In 24 cases (16.2%) discrepancies were due to a laboratory error. An explanation was found for the discrepant patterns in 119 cases (2 different strains, etc.) while no conclusion could be drawn in 5 cases.