During the past 10 years, there has been an increasing interest in the study of rapid-eye-movement (REM) sleep in neurodegenerative diseases and more particulary in Parkinson's disease (PD). This interest is justified by the strong association observed between these diseases and REM sleep behavior disorder (RBD). In the first section of this paper, a critical review of the literature on the presence of REM sleep disorders in PD is presented. Studies that show an association between PD and RBD are reviewed. Studies that report the presence of other REM sleep disorders in PD (short latency, abnormal length and/or proportion of REM sleep, increasing occurrence of hallucinations) are then discussed. Limitations of the criteria proposed by Rechtschaffen et Kales (1968) for the quantification of REM sleep are also presented. Some authors believe that dopaminergic (DA) agents used in the treatment of PD (levodopa, bromocriptine, pergolide, pramipexole and selegiline) could be a responsable factor for the occurence of REM sleep disorders observed in this disease. The literature concerning the impact of these DA agents on human REM sleep is therefore critically reviewed. It is concluded that DA agents cannot explain on their own the presence of REM sleep disorders in PD. Other causes, among which the disturbance of some neurochemical systems linked to the neuropathological process of the disease, must be considered in order to explain these REM sleep disorders. In the second section of this paper, we present the different pathophysiological hypotheses proposed to explain REM sleep disorders in PD, such as a dysfunction of the cholinergic, noradrenergic, serotonergic, dopaminergic or GABAergic neurons. Emphasis is placed on the role of cholinergic neurons of the pedunculopontine and laterodorsal tegmental nuclei, structures shown to be particularly impaired in PD. Neurophysiological, neuroanatomical and neuropharmacological studies demonstrate that these neurons are strongly implicated in the different REM sleep parameters (muscular atonia, electroencephalographic desynchronisation, ponto-geniculo-occipital spikes). Finally, future research directions are proposed.