The lack of standardization in chlamydia serology has made interpretation of published data difficult. This study was initiated to determine the extent of interlaboratory variation of microimmunofluorescence (MIF) test results for the serodiagnosis of Chlamydia pneumoniae infections. Identical panels of 22 sera were sent to 14 laboratories in eight countries for the determination of IgG and IgM antibodies by MIF. Although there was extensive variation in the numeric titer values, the overall percentage agreement with the reference standard titers from the University of Washington was 80%. For results by serodiagnostic category, the best agreement was for four-fold rise in IgG titers, while the lowest agreement was for negative or low IgG titers. Agreement for IgM titers was 50%-95%. Four laboratories failed to discern false-positive IgM titers possibly because of the presence of rheumatoid factor. Further studies are underway to determine the source of interlaboratory variation for the MIF test.