Fully oxidized D-neopterin in serum can be measured by HPLC. Serum samples were preincubated with ferric nitrate/EDTA solution to remove any dihydroneopterin, which is unstable and may give spuriously high results because of its conversion to D-neopterin. Pterins were extracted onto solid-phase propylbenzenesulfonic acid minicolumns and eluted with a 1:5 (by vol) mixture of ammonia solution (308 g/L) in acetonitrile. Extracts were evaporated and then reconstituted in mobile phase (50 mL of methanol per liter of 50 mmol/L phosphate buffer, pH 6.2) before injection. Separation was performed with a 25-cm ODS2 column (particle size, 5 microns) at 32 degrees C with fluorescence detection (lambda ex 360 nm, lambda em 440 nm). The between-batch CV was 7.1% and 5.6% for neopterin concentrations of 21.7 and 67.3 nmol/L, respectively. The limit of detection was 0.75 nmol/L, and the mean recovery of the extraction procedure was 90% for neopterin and internal standard. Correlation with a radioimmunoassay (x) gave y = 0.99x + 0.64 (r = 0.970, Sy/x = 2.75). The method allows daily analysis of serum D-neopterin in small batches and is currently used to monitor patients undergoing bone-marrow transplant.