OBJECTIVE To show the extent of diurnal variations in serum urate concentration and the possible consequences on the clinical interpretation. METHODS Nineteen controls and 40 patients took part in this study. Preliminary investigations on selected subjects provided blood samples at regular intervals between the hours of 0800 and 2300 hours. From these observations the timing of venepuncture for the remaining subjects was chosen at 0800-0900 hours, with a second sample taken between 1700 and 1800 hours. Analytical quality was monitored using quality control procedures and all samples were analysed in a single batch to minimise errors. RESULTS All subjects, except some hypertensive elderly women, showed a diurnal rhythm. The mean serum urate value between 0800-0900 hours was higher than that observed between 1700-1800 hours. A decrease of up to 30% was observed in a group of diabetic patients. Nine subjects in the study had hyperuricaemia in the morning, but only six in the afternoon, indicating the requirement of a reference interval that takes account of diurnal variations. CONCLUSIONS When studying the temporal relation for the medical importance, or when attempting to interpret the serum urate concentration, it is important to consider the diurnal rhythms of this analyte.