Many examples of biased sex ratios are known in natural populations of plants. Proximal causes of these biases are gender diphasy (sex changing), differential mortality between male and female genets, differential rates of clonal growth (numbers of ramets per genet), and differential flowering (differences in flowering frequency or age to maturity). In the western North American shrub Oemleria cerasiformis we determined sex ratios for 60 natural populations and found an excess of males in 56 populations. The male bias was greatest in populations with little recent recruitment. Sampling of young and old plants indicated that males flowered at an earlier age than females, which led to a transient flowering bias in very young plants, and that the genet sex ratio was 1:1 in young mature plants but male biased in old plants, as a result of differential mortality. Examination of dead genets confirmed that mature females have higher mortality rates. Females also have greater reproductive effort and slower growth rates than males. The major cause of biased sex ratios in O. cerasiformis is greater mortality of female plants during their reproductive years, which probably arises directly or indirectly from their greater allocation to reproduction.