A nondipping blood pressure (BP) pattern is common in patients with obstructive sleep apnea (OSA). However, it is unclear how useful a nondipping BP pattern is in screening for OSA. In this cross-sectional study, we recruited consecutive patients with clinical indications for performing ambulatory BP monitoring evaluating the following dipping patterns: (1) normal: ≥10% but <20%; (2) extreme: ≥20%; (3) reduced: ≥0% but <10%; and (4) reverse (riser): <0%. Sleep questionnaires and sleep studies were performed within 7 days after ambulatory BP monitoring. OSA was defined as an apnea-hypopnea index ≥15 events/h. We evaluated 153 patients (OSA frequency, 50.3%). Patients with OSA had higher BPs during sleep, were taking more antihypertensive drugs, and more frequently used hypertensive drugs during the night than patients without OSA. Considering systolic BP, the frequency of OSA in patients with reverse dippers (73.5%) was higher than normal (37.3%), extreme (46.2%), and reduced dippers (49.1%; P =0.012). For diastolic BP, OSA was more common in reduced (66.7%) and reverse dippers (69.6%) as compared to normal (41.4%) or extreme dippers (33.3%; P =0.007). In the regression analysis, reverse systolic dipper was independently associated with OSA (odds ratio, 3.92; 95% CI, 1.31–11.78). Both reduced and reverse diastolic dippers increased the likelihood of OSA for 2.7-fold and 3.5-fold, respectively. Snoring and positive sleep questionnaire findings were associated with a modest increase in the accuracy of reverse dipping pattern for predicting OSA. In conclusion, reverse systolic, as well as reduced and reverse diastolic dippers are independently associated with OSA.