Disentangling the genetic and environmental influences of gambling is important for explaining the roots of individual differences in gambling behavior and providing guidance for precaution and intervention, but we are unaware of any comprehensive and systematic quantitative meta-analysis. We systematically identified 18 twin studies on gambling in the meta-analysis. The correlation coefficients within monozygotic (MZ) and dizygotic (DZ) twins, along with the corresponding sample size, were used to calculate the proportion of the total variance accounted for by additive genes (A), dominant genes (D), the shared environment (C), and the non-shared environment plus measurement error (E). We further assessed the moderating effects of gambling assessment (symptom oriented assessment vs. behavior oriented assessment), age, and sex. The whole sample analyses showed moderate additive genetic (a2 = 0.50) and non-shared environmental influences (e2 = 0.50) on gambling. The magnitude of the genetic influence (a2) was higher for disordered gambling assessed with symptom oriented assessment (53%) than for general gambling assessed with behavior oriented assessment (41%). Additionally, the magnitude of the genetic influence (a2) was higher for adults (53%) than adolescents (42%). Genetic influence (a2) was greater for male (47%) gambling than female (28%) gambling. Shared environment had noticeable effects on female gambling (c2 = 14%) but zero effect on male gambling. In conclusion, gambling behavior was moderately heritable and moderately influenced by non-shared environmental factors. Gambling assessment, age, and sex significantly moderated the magnitude of genetic and environmental influences on gambling. Note that the number of studies might serve as a limitation.