[en] In the radioenzymatic assay of catecholamines, using catechol-O-methyltransferase, the yield of labelled product is frequently less than the expected value. This has been attributed by some workers to losses during the periodate oxidation of the O-methylated derivatives. Using spectrophotometric methods, we have demonstrated that there is no oxidated demethylation of the methoxycatecholamines during periodate oxidation. In a novel technique designed to determine the specific activity of small quantities of tritiated S-adenosylmethionine, high performance liquid chromatography with electrochemical detection was used to measure the amount of adrenaline formed from unlabelled noradrenaline, in the presence of phenyl-ethanolamine-N-methyltransferase and tritiated S-adenosylmethionine. As well as the real losses occurring during solvent extraction and thin layer chromatography (demonstrated spectrophotometrically), apparent 'loss' of radiolabelled product may be due to the assumption that the value for the specific activity of the tritiated S-adenosylmethionine and the determination of the product's radioactivity are both absolutely accurate, with no allowance being made for the normal and expected experimental errors in such measurements. (Auth.)