The amount of night-to-night variability in sleep apnea (SA) and sleep-related periodic leg movements (PLMs) is largely unknown but, despite this, clinical decisions are based on single-night studies in many clinical sleep laboratories. We examined variability in SA and PLMs over three nights in 46 community-resident seniors. No evidence was found for either a first-night effect or a directional trend across nights in either the Respiratory Disturbance Index (RDI) or the Movement Index (MI), despite a prominent first-night effect on pattern of sleep. Duration of apneas/hypopneas and degree of associated heart rate change and oxygen desaturation in subjects with SA and intermovement interval in subjects with PLMs also failed to show systematic change across nights. However, if a cut-off score of 5/h for RDI and MI was used, the classification recorded on the first night did differ from the classification given on at least one of the other nights in 43% of the subjects. The magnitude of fluctuation in RDI or MI from night to night was large enough in some subjects that, in a clinical situation, decisions based on one night would have been entirely different had the subject been studied on a different night. Night-to-night variability in RDI and MI within subjects also was associated with significant alterations in the sleep pattern. We conclude that caution should be taken in drawing conclusions from single-night studies, especially in individuals with relatively mild forms of SA and PLMs where nightly variations could easily place them above or below an arbitrary cut-off score.