We describe here results from the United Kingdom National External Quality Assessment Scheme (UK NEQAS) Thrombophilia Screening Program, in which an average of 21% of 280 centers reported an incorrect diagnosis for a series of plasma samples. Three case studies are described, showing causes of error in individual laboratories, related to the source of reference plasma or reagents. Methodological bias is also described. For protein C (PC) assays 18% of centers reported PC deficiency in a patient homozygous for factor V Leiden. Studies in the NEQAS laboratory confirmed the effect of activated protein C resistance (APCR) on clot-based PC activity assays. Differences in results obtained for PS-deficient subjects with different protein S (PS) activity kits are reported; several subjects would be misdiagnosed as normal with one kit if the manufacturer's reported reference range was adopted instead of a locally determined reference range. Antithrombin (AT) assays were shown to vary in their sensitivity to different molecular defects in the antithrombin gene; 77% of centers employing human thrombin-based activity assays reported a normal AT level in a patient with antithrombin Cambridge II. Sensitivity of the APC resistance test in the absence of factor V-deficient plasma was shown to be improved through normalization of results, and errors in the genetic diagnosis of factor V Leiden and the P20210A prothrombin gene mutation are described. Errors in the diagnosis of thrombophilic defects can therefore be identified through participation in EQA programs, and following dissemination of information, improvements in diagnosis can be demonstrated.