Introduction: Patients with complex sleep apnoea (CompSAS) have obstructive sleep apnoea and experience persistent central apnoeas when exposed to positive airway pressure. Elevated loop gain is one of the postulated mechanisms of CompSAS. We speculated that bilevel positive airway pressure—spontaneous (BPAP-S), by producing relative hyperventilation, may more readily produce CompSAS activity than continuous positive airway pressure (CPAP). If found to do so, a trial of BPAP-S might be a simple way of identifying patients with elevated loop gain who are at risk for CompSAS. Materials and methods: Thirty-nine patients with complex sleep apnoea were included in the study. Segments of NREM sleep on CPAP and BPAP-S matched for body position and expiratory airway pressure (“comparison pressure”) were retrospectively analysed. Correlations between clinical and demographic variables and polysomnographic response to CPAP and BPAP-S were sought. Results: There was no difference in any of the polysomnographic indices on CPAP and BPAP-S. In 19 patients the use of CPAP was associated with lower AHI at the comparison pressure; in 20 patients the opposite was true. No clinical variables correlated to the differential response to CPAP vs. BPAP-S. Conclusions: BPAP-S was not more effective than CPAP in stimulating complex sleep apnoea activity.