Sex allocation theory is often able to make clear predictions about when individuals should facultatively adjust their offspring sex ratio (proportion male) in response to local conditions, but not the consequences for the overall population sex ratio. A notable exception to this is in sex changing organisms, where theory predicts that: (1) organisms should have a sex ratio biased toward the "first" sex: (2) the bias should be less extreme in partially sex changing organisms, where a proportion of the "second" sex matures directly from the juvenile stage; and (3) the sex ratio should be more biased in protogynous (female first) than in protandrous (male first) species. We tested these predictions with a comparative study using data from 121 sex changing animal species spanning five phyla, covering fish, arthropods, echinoderms, molluscs, and annelid worms. We found support for the first and third predictions across all species. The second prediction was supported within the protogynous species (mainly fish), but not the protandrous species (mainly invertebrates).