1.Closely related taxa may be maintained as distinct species by a variety of reproductive isolating mechanisms. These include: inability to produce hybrid offspring, endogenous selection against hybrids in the form of genomic incompatibilities, and exogenous selection observable in genotypeby-environment interactions. To understand the relative importance of these three isolating mechanisms, we performed hand-pollination and reciprocal transplant experiments in a natural plant hybrid zone. 2. We measured reproductive isolation by making crosses between two parent species of Penstemon and naturally occurring hybrids. Inclusion of reciprocal hybrid crosses allowed us to determine whether fitness components differed depending on the identity of the mother. 3. Hybrid performance was evaluated in the greenhouse and in a reciprocal transplant experiment in the field. We measured fruit set, seed set, seed weight, time required for fruits to mature and seedling growth for potted plants. To test for exogenous isolation, we planted pure parents, reciprocal F 1 hybrids and later generation hybrids in a reciprocal transplant experiment, and measured survival. 4. On average, interspecific crosses produced as many seeds as conspecific crosses. Hybrid performance was also equal to or greater than parents in all environments, including the greenhouse and all field gardens, indicating a lack of endogenous isolation. Parent species and reciprocal F 1 hybrids differed in many traits measured. In each field garden, the hybrid with the native cytoplasm had a higher survival rate, suggesting local adaptation to different elevations. 5. Synthesis . Exogenous factors that differ along elevational gradients can be more important than intrinsic genetic incompatibilities in determining the fitness of plant hybrids. Our results illustrate the importance of studying hybrid performance in multiple environments and in generating reciprocal hybrids to test for isolating mechanisms in natural hybrid zones.