In the Netherlands the anal. performance of apo A-I and B determinations was evaluated. Performance was tested by the Dutch EQAS (SKML) using value-assigned, commutable control materials. Out of 150 participating clin. chem. laboratories, 22 laboratories routinely perform apo A-I and B analyses. Pooled data from six two-monthly surveys in 2004 revealed a mean within-lab precision for apo B and apo A-I of 4.5% and 4.0%, resp. Mean absolute bias for apo B and apo A-I was 6.8% and 8.6%, resp. Mean interlaboratory precision for apo B and apo A-I was 8.6% and 9.9%, resp. One year after IVD 98/79/EC implementation, average within lab precision and average absolute bias of apo A-I and B applications do not meet desirable performance criteria based on biol. variation. Notwithstanding the availability of internationally recognized reference systems and legal support, professional and EQAS support are needed to improve apo A-I and B standardization.