OROS (melatonin), an oral osmotic system for controlled drug delivery, was evaluated in an open-label, two-way crossover study to test the feasibility of continuous overnight melatonin delivery. Twelve healthy subjects with no sleep disorders, ranging from 60 to 73 years of age, were enrolled in the study. Two doses of melatonin (1 x 110 micrograms and 4 x 110 micrograms) were administered on two separate occasions. Endogenous baseline nighttime serum melatonin concentrations were measured the night before each treatment. Following treatment at 2100 hours, the lights were extinguished at 2200 hours and remained so, except during blood sample collection, which was performed under dim light (< 50 lux) at specified times. Serum samples were analyzed for melatonin by an LC/MS/MS method. In addition, safety measurements such as vitals and serum samples for endocrine functions were measured both prior to and after melatonin dosing. The serum melatonin concentration profile following OROS (melatonin) dosing mimicked the normal endogenous serum melatonin concentration-time profile. The mean maximal melatonin concentration occurred at 3 a.m. The mean AUCs of endogenous melatonin before the two treatment days were 248 and 234 pg.h/mL, respectively. Serum concentrations of melatonin corrected for endogenous production increased proportionally with dose, with AUCs of 288 and 1069 pg.h/mL, respectively. Deconvolution of the serum concentration data showed good correlation between the in vitro amount released and the in vivo amount absorbed, suggesting continuous absorption throughout the gastrointestinal tract. Less than 5% residual content was observed in the recovered OROS system. Minimal changes in serum hormone concentrations (luteinizing hormone, follicular stimulating hormone, and prolactin) and no serious adverse events were observed following OROS treatment in these subjects. Delivery of melatonin with OROS formulation may result in a physiologic nocturnal profile in elderly subjects.