Study objectives: To determine the effect of transvenous phrenic nerve stimulation (TPNS) on the composition of the nocturnal hypoxemic burden in patients with CSA. Methods: We analysed oximetry data from baseline and follow-up overnight polysomnograms (PSG) in 134 CSA patients with implanted TPNS randomised (1:1) to neurostimulation (treatment group; TPNS on) or no stimulation (control group; TPNS off) from the remedē System Pivotal Trial. The hypoxemic burden was quantified using a battery of metrics, including the oxygen desaturation index (ODI), the relative sleep time spent below 90% SpO2 (T90) due to acute episodic desaturations (T90_desat) and due to non-specific and non-cyclic drifts of SpO2 (T90_non-specific). Mean change from baseline is provided. Results: TPNS titrated to reduce respiratory events significantly reduced the ODI in the treatment group by −15.85 h⁻¹ ± 1.99 compared to the control group, which increased 1.32 h⁻¹ ± 1.85 (p 〈0001) and shortened the relative T90 duration by −3.81 percentage points ± 1.23 vs. 0.49 percentage points ± 1.14 increase (p = 0.012). This shortening of T90 was primarily accomplished by reducing the brief cyclic desaturations (T90_desaturation: −4.32 percentage points ± 0.98 vs. 0.52 percentage points ± 0.91, p = 0.0004) while notable non-specific drifts in SpO₂ remained unchanged (T90_non-specific: 0.18 percentage points ± 0.62 vs. -0.13 percentage points ± 0.57, p = 0.72). Conclusions: TPNS appears to significantly reduce the nocturnal hypoxemic burden due to sleep-disordered breathing, but a considerable nocturnal hypoxemic burden from other sources remains. Further investigations are warranted to identify the best strategy to reduce the nocturnal hypoxemic burden beyond preventing respiratory events.