Objective: Continuous Positive Airway Pressure (CPAP) is the main therapy for obstructive sleep apnea (OSA); nevertheless, uncertainty remains about the effectiveness of CPAP not only in controlling OSA-related hypertension, but also in improving the metabolic dysregulation that characterizes OSA patients. This meta-analysis of randomized controlled trials (RCTs) aimed to investigate whether CPAP, compared to other control treatments (sham CPAP, oral placebo or standard care), could improve glucose metabolism (fasting plasma glucose [FPG] or glycated haemoglobin or fasting plasma insulin [FPI] or homeostasis model assessment [HOMA] of insulin resistance [IR]) and/or lipid metabolism (total, HDL, LDL-cholesterol or triglycerides) in OSA patients. Design and method: Relevant papers were searched in three different databases (MEDLINE, EMBASE and Web of Science) from inception to 31st Dec 2018 through specific search terms and selection criteria. Results: From 2,289 articles, 26 RCTs were included. Regarding glycemic metabolism, neither FPG nor glycated haemoglobin were reduced by CPAP. However, patients on CPAP showed an improvement in insulin sensitivity, as witnessed by a significant reduction of FPI (Standardized difference in means [SDM] = 0.205 [95%CI, 0.07–0.339], p = 0.003) and HOMA IR (SDM = 0.176 [95%CI, 0.056–0.296], p = 0.004). In sensitivity analysis the favourable effects of CPAP were maintained in prediabetic/type 2 diabetic patients and in those presenting an apnoea-hypopnea index > 30 events/h. Regarding lipid metabolism, OSA patients treated with CPAP showed a significant reduction in total cholesterol (TC; SDM = 0.138 [95%CI, 0.034–0.242], p = 0.009). In subgroups analysis, greater adherence to CPAP along with severe nocturnal oxygen desaturations at baseline (SpO2-nadir < 77%) was associated with a significant decrease of TC. Furthermore, dividing the primary studies according to the median TC at baseline (198 mg/dL), the positive effects of CPAP were maintained only in those that presented a TC above the median. Conversely, CPAP did not modify triglycerides, HDL- and LDL-cholesterol. Conclusions: CPAP treatment significantly improves insulin sensitivity and reduces TC in OSA patients but with a low effect size. The effectiveness of CPAP is higher in patients with greater CPAP usage, and with a greater number of apnoeic events and oxygen desaturations at baseline. Concerning TC, patients presenting higher levels at baseline benefit the most from CPAP.