External quality assessment is a tool to compare the result of a particular laboratory test in relation to those of other laboratories as well as to assess the performance of a laboratory test over a prolonged period of time. We evaluated the relationship between the between-laboratory variation and the sample category (normal, borderline, and abnormal) for antithrombin, protein C, protein S, and the activated protein C resistance test. Only for antithrombin and protein S was a significant relationship (0.004 < p < 0.012) observed. The effect of the between-laboratory variation of the different sample categories on the clinical interpretation was investigated. With the exception of free protein S antigen, all variables showed a significant relationship (0.004 < p < 0.045) between the sample category and the percentage of misclassification. Because in clinical practice a stable test performance over a prolonged period of time is important, we evaluated the quality of test performance using the long-term analytical coefficient of variation (LCVa). A wide range in the LCVa was observed for antithrombin, protein C, and protein S. Less than half of the participants could fulfill the quality specification for diagnostic testing (LCVa < or = 0.58 x total biological variation). This study shows that a more stable performance of laboratory tests involved in the screening of thrombophilia over a prolonged period of time is necessary.