OBJECTIVES: Short sleep duration and poor sleep quality in children have been associated with concentration, problem behavior, and emotional instability, but recently also with disrupted autonomic nervous function, which predicts cardiovascular health. Heart rate variability (HRV) was used as noninvasive indicator of autonomic function to examine the influence of sleep. DESIGN: Cross-sectional and longitudinal observational study on the effect of sleep on HRV. PARTICIPANTS: Belgian children (5-11 years) of the ChiBS study in 2010 (N = 334) and 2011 (N = 293). INTERVENTIONS: N/A. METHODS: Sleep duration was reported and in a subgroup sleep quality (efficiency, latency, awakenings) was measured with accelerometry. High-frequency (HF) power and autonomic balance (LF/HF) were calculated on supine 5-minute HRV measurements. Stress was measured by emotion and problem behavior questionnaires. Sleep duration and quality were used as HRV predictors in corrected cross-sectional and longitudinal regressions. Stress was tested as mediator (intermediate pathway) or moderator (interaction) in sleep-HRV associations. RESULTS: In both cross-sectional and longitudinal analyses, long sleep latency could predict lower HF (parasympathetic activity), while nocturnal awakenings, sleep latency, low sleep efficiency, and low corrected sleep duration were related to higher LF/HF (sympathetic/parasympathetic balance). Parental reported sleep duration was not associated with HRV. The significances remained after correction for stress. Stress was not a mediator, but a moderator (enhancer) in the relationship between sleep quality and HRV. CONCLUSIONS: Low sleep quality but not parent-reported low sleep duration leads to an unhealthier heart rate variability pattern (sympathetic over parasympathetic dominance). This stresses the importance of good sleep quality for cardiovascular health in children.