In contrast to rheumatoid arthritis (RA), no systematic investigation of diurnal variation has been carried out in polymyalgia rheumatica (PMR). The aim of the study was to provide the often-requested documentation of the 24-h time course of clinical symptoms in PMR and relate them to concentrations during the day of melatonin, inflammatory cytokines, and cortisol. Furthermore, the effects of 14 days of prednisolone treatment were studied. Ten glucocorticoid-naive patients newly diagnosed with PMR and seven non-PMR control subjects were studied for 24 h before treatment and during the 14th day of treatment with 20 mg/day of prednisolone. Global pain and generalized muscle stiffness were monitored by using visual analogue scales, and blood was drawn repeatedly. In untreated patients, pain and stiffness peaked in the early morning, showing a plateau between 04:00 and 08:00, and then declined to a nadir at 16:00 (2P < 0.05). Plasma concentrations of interleukin (IL)-6, IL-8, tumor necrosis factor (TNF)-α, IL-1β, and IL-4 varied with time in both groups (2P < 0.05) and peaked between 04:00 and 08:00. Furthermore, except for IL-1β, concentrations of these cytokines and of IL-10 were higher throughout the 24-h observation period in patients than in control subjects (2P < 0.05). Also, melatonin and cortisol were consistently higher in patients (2P < 0.05) and varied with time (2P < 0.05), peaking around 02:00 and 08:00, respectively. In patients, prednisolone abolished symptoms, normalized C-reactive protein, and reduced melatonin, IL-6, IL-8, and TNF-α concentrations (2P < 0.05), while IL-10 increased between 10:00 and 14:00. In PMR, key symptoms show diurnal variation. Furthermore, in PMR, concentrations of melatonin, several pro- and anti-inflammatory cytokines, and cortisol are increased throughout the day and show diurnal variation, as also seen in healthy subjects. The time courses and the inhibitory effects of prednisolone indicate that in PMR, as proposed for RA, melatonin stimulates cytokine production, which in turn accounts at least partly for the symptoms. Furthermore, overall, cortisol may downregulate cytokine production and symptoms. Stimulation of IL-10 secretion may participate in the anti-inflammatory effects of prednisolone. These findings support use of chronotherapy in PMR and encourage study of circadian variations in other inflammatory autoimmune diseases.