Insomnia is prevalent among patients with alcohol use disorder (AUD), potentially undermining treatment and increasing the risk of relapse. Cognitive behavioral therapy for insomnia (CBT-I) is the recommended first-line treatment for insomnia, but its efficacy is not well-characterized in patients across the spectrum of AUD. The aim of this meta-analysis was to quantify the effectiveness of CBT-I in improving insomnia severity and alcohol-related outcomes in adults with heavy alcohol use and/or varying levels of AUD severity and comorbid insomnia. MEDLINE, Cochrane Library, PsycINFO and ClinicalTrials.gov were systematically searched (up to February 2024) to retrieve randomized controlled trials (RCTs). Multilevel meta-analyses were conducted to estimate mean differences over time in insomnia severity, measured using the Insomnia Severity Index (ISI), as well as in alcohol craving and alcohol-related psychosocial problems between CBT-I and control groups. For the number of heavy-drinking/abstinent days, incidence rate ratios were estimated. Risk of bias was assessed using the Risk of Bias 2 (RoB 2) tool. Eight RCTs encompassing 426 adults (68.78 % men) were included. Compared with control conditions, CBT-I resulted in a large reduction of insomnia severity post-treatment [estimated ISI reduction = -5.51, 95% CI (-7.13 to -3.90)], which was maintained at 1-to-3-month [7 studies; estimate = -4.39, 95% CI (-6.08 to -2.70)], and 6-month follow-up [4 studies; estimate = -4.55, 95% CI (-6.77 to -2.33)]. Alcohol-related outcomes were reported less consistently, and no significant differences were found. The included trials were judged to have a low or moderate overall risk of bias for the assessment of all outcomes. CBT-I effectively reduces insomnia severity across the spectrum of AUD, supporting wide implementation in AUD prevention and treatment settings. PROSPERO REGISTRATION NUMBER: CRD42023464612.