The credibility of evidence of various environmental risk factors for alcohol and tobacco use disorders (AUD/TUD) needs to be graded to identify groups to target with selective prevention. A systematic umbrella review was conducted (PubMed/PsycINFO), grading credibility of meta-analyses of prospective/retrospective observational cohort studies assessing risk/protective factors for AUD/TUD, applying established quantitative criteria. Sensitivity analyses were conducted. Quality of eligible meta-analyses was assessed with AMSTAR-2. Out of 8464 unique references, 80 full text articles were scrutinized, and 12 meta-analyses, corresponding to 21 individual estimates of 12 putative risk/protective factors (n = 241,300), were included. In main analyses no association had convincing nor highly suggestive evidence for AUD/TUD. Six associations had suggestive evidence for AUD, two for TUD. Among these, in sensitivity analyses without >1000 cases criterion, convincing evidence emerged for parental alcohol supply, and impulsivity traits in college students for AUD, and attention-deficit/hyperactivity disorder for TUD. Other associations were supported by weak evidence/were not nominally significant. Few risk factors identified at-risk groups where selective preventative strategies could be developed to prevent AUD/TUD.