The diverse features of floral morphology are often thought to be well-designed mechanisms to manipulate plant mating systems. We evaluated the effectiveness of one such mechanism, anther-stigma separation (herkogamy), in controlling variation in the level of outcrossing among plants in a population of Gilia achilleifolia. Variation in outcrossing rates within populations has the potential to influence the coevolution between inbreeding depression and mating system. Using four polymorphic allozymes, we compared the outcrossing-rate estimates of two groups of individuals under natural conditions: one group with low herkogamy and another with high herkogamy. The high herkogamy group had a higher outcrossing rate (0.572) than the low herkogamy group (0.335). This suggests that the within-population variation in outcrossing rate could potentially cause the previously observed association between herkogamy and inbreeding depression (Takebayashi and Delph, 2000). A previous study of floral traits among G. achilleifolia populations failed to detect a relationship between herkogamy and outcrossing rate, demonstrating that the functionality of traits may be obscured in among-population studies as a consequence of uncontrolled environmental variation. Additionally, the effect of herkogamy on outcrossing rate in delayed selfers such as G. achilleifolia may be particularly prominent when pollinator availability is low. Our population-level estimate of outcrossing rate (0.444) was somewhat lower than an estimate from the same population, 15 years prior to our study (0.75), suggesting that pollinator availability may fluctuate among years. Both within-year and among-year variation in the outcrossing rate may have a strong influence on mating-system evolution.