When more pollinator visits are required to remove pollen than are required to fertilize all of the ovules, male-functioning plants are expected under theory to allocate more resources to secondary floral structures and products. We found that for the gynodioecious shrub, Hebe stricta, there were no differences between the perfect-flowered ("male") morph and the male-sterile ("female") morph in many important aspects of their displays (number of flowering shoots, number of flowers per inflorescence, timing of floral opening, size of corolla), but the flowers on the male morph had significantly larger anthers that extended a greater distance out of the floral tube. Flowers on males produced over four times as much nectar, but the standing crop of nectar did not differ between the two sexual morphs in nature. Overall, pollinators preferred to visit flowers on male plants, but different species of pollinators responded differently to the two plant morphs. Finally, a pollen-addition experiment demonstrated that fruit set on the females was not significantly limited by pollen; hence, the fewer visits by pollinators to this morph seems sufficient to insure maximal fruit set given the resources available (satisfying Bateman's principle). We suggest that the greater number of visits to males is in response to their pollen production and to their greater nectar production, but that an ideal free distribution is set up in nature, which insures a sufficient number of visits to the females. These results, in combination with data on nectar production of cosexual and gynodioecious species drawn from the literature, support the conclusion that fitness through male function is enhanced by allocating more resources to attract pollinators than is needed for maximal fitness through female function.