In spite of the increasing interest in the analysis of methylmalonic acid and total homocysteine in plasma, data on interlaboratory variation is lacking. We report the results of an external quality assessment study with the participation of 15 laboratories in the Scandinavian countries performing these analyses on a regular basis. For methylmalonic acid, using serum, heparin fluoride plasma and an aqueous sample, CVs were found in the range of 11-17%. For total homocysteine using EDTA plasma, heparin fluoride plasma and an aqueous sample, CVs were in the range of 6-12%. For both analytes, a significant correlation between the individual recoveries of added analyte and the results for the aqueous sample was found, suggesting that the use of inconsistent calibrations in the participating laboratories are contributing to the interlaboratory variation. An acceptable range for results from the individual laboratory was calculated using data on the biological within-subject and between-subject variations reported in the literature. These ranges were violated by several laboratories when using the consensus mean or median as target values. Even if the results of the present study document a reasonable standard in the measurement of methylmalonic acid and total homocysteine in plasma in the participating laboratories there is room for improvement and a permanent scheme of external quality assessment using relevant samples is essential. From 1997, a regular scheme has been available from our laboratory.