The pharmacokinetics and relative bioavailability of cyclobarbital calcium have been studied after oral administration of Phanodorm, of tablets according to the Formularium Nederlandse Apothekers (1968; FNA), and an aqueous solution. Six healthy volunteers participated in the investigation on three occasions and each received the three preparations. The dose administered was 300 mg cyclobarbital calcium. Plasm concentrations of cyclobarbital were determined at regular intervals. Absorption from the three preparations was rapid was faster from the solution. Peak concentrations were usually attained within 1 h. The eleimination of cyclobarbital could be described by a single first-order process with an average half-life of 11.6 h (range 8 - 17 h). There was little intra-subject variation of the half-life. Relative bioavailability for each volunteer was estimated by comparing the areas under the plasma concentration curves. The RNA-tablets and Phanodorm exhibited similar bioavailability, whereas the average bioavailability of the solution was 78% of that of FNA-tablets; the reason for this unexpected finding is unknown. It was concluded that cyclobarbital cannot be regarded as a uniformly suitable drug for the treatment of insomnia. The long half-life that was apparent in some of the volunteers (15 - 17 h) creates a substantial risk of residual effects on the following morning. In principle, however, the calcium salt of cyclobarbital may be used for induction of sleep, because of its rapid absorption.