The Geschwind-Behan-Galaburda and sexual differentiation models predict an association between elevated foetal androgen exposure and left-handedness whereas the callosal hypothesis predicts the opposite. We present a meta-analysis of correlations between handedness and digit ratio (2D:4D), a putative marker of prenatal testosterone. Left-handedness predicted low (male-typical) right-hand digit ratio (R2D:4D), high (female-typical) left-hand digit ratio (L2D:4D), and low R2D:4D-L2D:4D directional asymmetry (D[R−L]). Effect sizes were extremely small and not moderated by sex or method of measuring handedness or 2D:4D. The same general pattern was observed after excluding the very large study (110,329 males, 90,412 females) of Manning and Peters (. Digit ratio (2D:4D) and hand preference for writing in the BBC Internet Study. Laterality: Asymmetries of Body, Brain and Cognition, 14(5), 528–540. doi:10.1080/13576500802637872); however, no significant effects for R2D:4D were observed once these samples were removed. The results do not confirm any theory linking prenatal androgens with handedness, so we speculate they instead reflect the mechanical action of writing causing subtle changes in the musculature and/or fat pads of the fingers. Gripping a pen/pencil might cause an increase in 2D relative to 4D (and/or decrease in 4D relative to 2D) resulting in higher ratios on the writing-hand; furthermore, this could differ between left- and right-handers due to writing in the left-to-right direction (as in English) having asymmetrical effects depending on which hand is used.