Objective To investigate the features and relative factors of blood pressure variability (BPV) in patients suffering from severe obstructive sleep apnea hypopnea syndrome (OSAHS) with and without hypertension. Methods A total of 125 patients with severe OSAHS in our hospital from August 2016 to December 2017 were divided into hypertensive group (n=64) and normotensive group (n=61). Blood pressure (BP) was continuously monitored and measured via pulse transit time (PTT) during polysomnography (PSG) and calibrated when 3 consecutive stable supine cuff values were obtained with patients in a conscious state. Variations in BP and pulse were incorporated in PSG. The amplitude of BP fluctuation (ΔBP) was reflected by the difference between the peak and bottom values of post apneic SBP recorded during an obstructive respiratory event, and the number of ΔBP>10 mmHg(1 mmHg=0.133 kPa)per hour of sleep time was used as indicator for frequency of significant BP fluctuations (BP index). To explore the effects of arousal and hypoxia on BP fluctuations, nocturnal respiratory events were classified into three types:type A with only oxygen desaturation; type B with arousal but without oxygen desaturation; and type C with both arousal and oxygen desaturation. △BP was compared between 3 types. SPSS statistics 16.0 was used for data analysis, and all data were expressed as mean±standard deviation. The independent sample Student′s t test was used for comparison between the groups and Chi-square test for comparison between the 3 types of apnea events. Pearson′s correlation and then a stepwise multiple regression analysis were performed to investigate the relationships between variables. Results Compared with the normotensive group, the hypertensive group had more severe sleep-disordered breathing (SDB) and significant higher ΔBP [(15.4±4.5) vs (10.9±2.6) mmHg] and BP index [(57.0±16.5) vs (22.7±12.0) times/h] (P＜0.001 in both). Analysis of the nocturnal respiratory events showed that ΔBP were remarkably greater following type A than type B in hypertensive group[(15.4±4.9) vs (11.9±3.8) mmHg] and normotensive group [(10.6±3.1) vs (9.5±2.4) mmHg] (P＜0.001 in both), and that ΔBP following type C were significantly higher than both type A and type B. A stepwise multiple regression equation demonstrated that the mean level of BP had a higher correlation with BP index [awake interval:r²=0.454, P<0.001; sleep interval:r²=0.470, P<0.001)] and the percentage of sleep time with pulse oxygen saturation (SpO₂) <90% (awake interval:r²=0.051, P<0.001; sleep interval:r²=0.073, P<0.001) than with other sleep disorder parameters. Conclusion In the patients with severe OSAHS, the occurrences of hypertension are closely associated with increased nocturnal blood pressure variability (BPV) and hypoxic duration and can be induced by both hypoxia and arousal, with hypoxia being a dominant factor.