In clinical practice, the seriousness of liver disease is assessed based on the combined information from clinical examination, routine biochemical tests, and liver histology. Recently, the assessment of hepatic lidocaine metabolism has been proposed as a quantitative liver function test offering valuable additional information. To evaluate whether this new liver function test reflects the combined clinical assessment, we prospectively measured lidocaine metabolism in 111 patients with well characterized liver disease. In addition, lidocaine test results were compared with the aminopyrine breath test and the galactose elimination capacity. Lidocaine (1 mg/kg) was injected i.v. and serum concentrations of its main metabolite monoethylglycinexylidide were determined after 15 min. The results varied widely and the means (+/- S.D.) were similar among patients with mild liver disease (46 +/- 23 ng/ml), but significantly (P < 0.05) lower among patients with Child class A cirrhosis (19 +/- 11 ng/ml) or Child class B or C cirrhosis (21 +/- 19 ng/ml). The [13C]aminopyrine breath test, however, gave a better discrimination among patients with increasing severity of liver disease than lidocaine metabolite formation. The galactose elimination capacity finally best separated patients with mild liver disease from those with cirrhosis. The correlations between any two of the different quantitative liver function tests were weak (R2 consistently < 0.2). We conclude that lidocaine metabolite formation, like other quantitative liver function tests that are based on the microsomal metabolism of model compounds, quantitates a very particular enzymatic reaction which may not be representative for the functional reserve of the entire organ.