Résumé : OBJECTIVE: Comparisons of sleep Slow Wave Activity (SWA) during successive sleep cycles rely on the assumption that SWA in a given cycle is independent of the number of ultradian cycles present in a night. This assumption was evaluated here. METHODS: Twenty-six healthy controls with no medical, sleep or psychiatric disorders were selected among 84 candidates and their sleep was recorded at home across 2 consecutive nights after two habituation nights. RESULTS: In comparison with nights with less cycles, nights with more cycles showed significantly more REMS but not more NREMS. No correlation was found between the number of cycles and the integrated SWA per night (epochs visually scored as NREMS). However, inverse correlations were found between the number of cycles and the SWA per cycle. This was significant on both nights in Cycle 1 and strong trends were found for the two subsequent cycles on Night 2. Comparable results were found after removal of nights containing suspected Skipped First REMS episodes. CONCLUSIONS: The SWA in a cycle was found to be inversely correlated to the number of cycles in the first 3 cycles in at least one of the two analyzed nights. SIGNIFICANCE: Differences in the number of cycles per night are a potential bias in the comparisons of SWA per cycles.