Mannitol is an osmotically active polyalcohol often present in fluids used for irrigation of exposed tissue during minimal invasive surgery. Since this polyol normally is not detected in human plasma to any significant extent, it may be used as a laboratory marker of absorption of mannitol-containing irrigative fluids during surgery. For this aim, we developed a photometric assay of mannitol in human blood or serum that may be performed in a near-patient setting. Following deproteinization of the sample with trichloroacetic acid, the supernatant is mixed with NAD+ and a commercially available preparation of mannitol 2-dehydrogenase and is incubated at pH 7.8 and at 37 degrees C for 30 to 60 minutes. At the end of the incubation period the solution is appropriately diluted and the concentration of NADH formed by oxidation of mannitol is determined photometrically at 340 nm. The limit of detection of serum mannitol with this assay is 0.05 mmol/l, the linear range of measurement extends to about 3 mmol/l. At analyte concentrations of 0.48, 1.38 and 3.48 mmol/l, coefficients of inter-assay variation of 12.1, 6.7 and 4.9%, respectively, were obtained. The analytical recovery of mannitol added to serum samples was close to 100%. Of 27 polyalcohols, monosaccharides and oligosaccharides tested, none exhibited a measurable substrate activity and only D-fructose significantly inhibited the oxidation of mannitol at sample concentrations above 10 mmol/l; the enzymatic reaction, however, was strongly affected by EDTA. The suitability of the assay as a routine diagnostic tool for detection and quantification of intraoperatively absorbed irrigation fluid was demonstrated by analyzing mannitol in serum samples obtained from 24 patients undergoing transurethral prostatectomy.