Sex ratios of flowering individuals in dioecious plant populations are often close to unity, or are male biased owing to gender-specific differences in flowering or mortality. Female-biased sex ratios, although infrequent, are often reported in species with heteromorphic sex chromosomes. Two main hypotheses have been proposed to account for female bias: (1) selective fertilization resulting from differential pollen-tube growth of female- versus male-determining microgametophytes (certation); (2) differences in the performance and viability of the sexes after parental investment. Here we investigate these hypotheses in Rumex nivalis (Polygonaceae), a European alpine herb with female-biased sex ratios in which females possess XX, and males XY₁Y₂, sex chromosomes. Using field surveys and a glasshouse experiment we investigated the relation between sex ratios and life-history stage in 18 populations from contrasting elevations and snowbed microsites and used a male-specific SCAR-marker to determine the sex of nonflowering individuals. Female bias among flowering individuals was one of the highest reported for populations of a dioecious species (mean female frequency = 0.87), but males increased in frequency at higher elevations and in the center of snowbeds. Female bias was also evident in nonflowering individuals (mean 0.78) and in seeds from open-pollinated flowers (mean 0.59). The female bias in seeds was weakly associated with the frequency of male flowering individuals in populations in the direction predicted when certation occurs. Under glasshouse conditions, females outperformed males at several life-history stages, although male seeds were heavier than female seeds. Poor performance of Y₁Y₂ gametophytes and male sporophytes in R. nivalis may be a consequence of the accumulation of deleterious mutations on Y-sex chromosomes.