Aims: To evaluate the potential interaction of 20 mg paroxetine and 1 mg alprazolam (early morning once-daily administration) on polysomnographic (PSG) sleep and subjective sleep and awakening quality, both after a single intake and after reaching a steady-state concentration. Methods: Twenty-two (11 for the PSG) healthy young volunteers of both sexes with no history of sleep disturbances (Pittsburgh Sleep Quality Index <5) participated in a double-blind, double-dummy, placebo-controlled, repeated-dose, 4-period, cross-over study. All volunteers received all 4 treatment sequences: paroxetine–alprazolam placebo (PAP); paroxetine placebo–alprazolam (PPA); paroxetine–alprazolam (PA), and paroxetine placebo–alprazolam placebo (PLA), in a randomized order. Each treatment was administered over 15 consecutive days, with a treatment-free interval of 7 days prior to the subsequent study period. In each experimental period, one PSG sleep study was performed on the 1st night (single-dose effects) and another study was performed on the 15th night (repeated-dose effects). Additionally, two other PSG studies were assessed: an adaptation recording, and a control night recording. All-night PSG recordings were obtained following standard procedures. Each 30-second period was scored according to the criteria of Rechtschaffen and Kales by means of an automatic sleep analysis system: Somnolyzer 24x7TM. A self-rating scale for sleep and awakening quality and early morning behavior was completed no later than 15 min after awakening over the 15 days of each experimental intervention. General lineal models (treatment/time) were applied separately to each variable. Results: (1) No significant effects were observed in any sleep variables when control nights were compared with the 1st night with PLA. (2) Sleep continuity: After PAP a clear awakening effect was seen both in the first and second evaluations, mainly in wake time, movement time, number of awakenings and stage-1 duration. After PPA an evident hypnotic effect was observed on night 1. This effect was mainly observed in maintenance variables and slightly in sleep initiation variables; it had decreased by night 15. After PA an intermediate behavior in the variables related to sleep continuity was seen, highlighting the absence of the tolerance phenomenon observed when PPA was administered alone. (3) Sleep architecture: The most important effects in REM sleep were observed after PAP; an increase in REM latency and decreases in REM sleep. PAP also induced decreases in the number of non-REM and REM periods and increases in the average duration of non-REM periods and sleep cycles. PA presented a similar pattern to PAP, and PPA similar to PLA. In relation to non-REM sleep, PA showed more stage-2 and less slow-wave sleep (SWS). (4) Subjective perception: No significant differences were observed between treatments while they were being taken, but impairments in subjective sleep quality, awaking quality, latency and efficiency were seen, mainly after PA but also after PPA discon...