We performed a reciprocal transplant experiment to estimate "parallel" adaptation to elevation and "unique" adaptation to local sites at the same elevation, using the frog Rana temporaria in the Swiss Alps. It is important to distinguish these two processes because they have different implications for population structure and ecological specialization. Larvae were reared from hatching to metamorphosis within enclosures installed in their pond of origin, in three foreign ponds at the same elevation, and in four ponds at different elevation (1500-2000 m higher or lower). There were two source populations from each elevation, and adults were held in a common environment for 1 year before they were crossed to produce offspring for the experiment. Fitness was a measure that integrated larval survival, development rate, and body size. Parallel adaptation to elevation was indicated by an advantage at the home elevation (11.5% fitness difference at low elevation and 47% at high elevation). This effect was stronger than that observed in most other studies, according to a survey of previous transplant experiments across elevation (N = 8 animal species and 71 plants). Unique local adaptation within elevational zones was only 0.3-0.7 times as strong as parallel adaptation, probably because gene flow is comparatively high among nearby wetlands at the same elevation. The home-elevation advantage may reduce gene flow across the elevational gradient and enable the evolution of habitat races specialized on elevation.