A main tenet of sex-allocation theory is that environmental stress should lead to increased maleness because reproducing through pollen is generally cheaper than producing fruits and seeds. Though this prediction has held for many species, it has been little tested for gynomonoecious plants, in which individuals produce both female and perfect flowers. We exposed eight ramets of each of 22 genets of a gynomonoecious goldenrod, Solidago altissima (Asteraceae), to a factorial combination of nutrient stress and herbivory by the gall-inducer Eurosta solidaginis (Tephritidae). Nutrient stress alone increased relative femaleness: Stressed ramets produced fewer flowers total and a higher ratio of ray (female) flowers to disk (perfect) flowers. Galling caused no change in fertilized ramets, but the combination of nutrient stress and galling caused an increase in relative maleness: Nutrient-stressed, galled ramets produced fewer flowers total and had a higher disk to ray ratio. In addition to being phenotypically plastic, floral-sex ratio had a great deal of genetic variation, with a broad-sense heritability of 0.68. While the floral-sex ratio responses of gynomonoecious plants may be more complicated than for plants of other breeding systems, they offer the potential to test and refine the already rich body of sex-allocation theory.