Human studies have reported inconsistent associations between the length ratio of the second finger to the fourth finger (2D:4D), which is a proxy for prenatal androgen load, and substance or computer use in adolescents and adults. This metaanalysis quantifies the magnitude of this relationship and investigates the roles of sex, definition of caseness, different forms of addiction, the hand measured (right hand versus left hand), and other cohort characteristics. Univariate random-effects meta-analyses were performed, and moderators were tested with Bonferroni-corrected meta-regression analyses. The study included 18 independent samples with a total of 175,955 participants (96,316 males and 79,639 females). There was a significant difference in 2D:4D between the substance and computer-using subjects and the controls for the combined sample (Hedge's g = − 0.178 [− 0.291; − 0.064]) and for males (Hedge's g = − 0.260 [− 0.399; − 0.122]), but not for females. These effects were amplified when only analyzing studies that compared dependent versus non-dependent subjects (combined sample: g = − 0.325 [− 0.492; − 0.157]; males: g = − 0.427 [− 0.564; − 0.291]), but did not reach significance in the subgroup of studies examining other parameters of substance and computer use. When analyzing different forms of substance and computer use separately, alcohol intake and computer use revealed a significant difference in the standardized mean. Again, the effects were amplified when analyzing the subgroup of males and the subgroup of studies comparing dependent versus non-dependent subjects, with effect sizes showing Hedge's g values as many as − 0.552 [− 0.785; − 0.319] (alcohol-dependent males). Thus, this meta-analysis confirms that lower 2D:4D is associated with substance and computer dependency. Further studies are encouraged to explore the link between intrauterine hormone environment and addiction risk.