Papilio canadensis encounters shorter, cooler summers in interior Alaska than in northern Michigan: average thermal sums are 583 vs. 985 Celsius degree—days (10°C base); mean daily temperature is 14.4 vs. 18.8°. The temperature physiology of P. canadensis could be evolutionarily conserved, or the species may be a composite of regionally adapted populations. We evaluated these hypotheses by comparing the developmental physiology of P. canadensis from Alaska and Michigan across a range of temperatures in the laboratory and field. Higher temperatures generally resulted in more rapid larval development, but the effects varied with insect population and host. At low temperatures (12°), Alaskan larvae grew faster than Michigan larvae (fifth instars doubled their fresh mass in 5.8 vs. 9.1 d), primarily due to 40% higher consumption rates. At high temperatures (30°), Alaskan larvae grew slower, faster, or the same as Michigan larvae, depending upon host. Effects of host quality were greatest at high temperatures. Elevated respiratory expenses in Alaskan larvae (35% higher than Michigan larvae) made them especially sensitive to host quality at high temperatures. Dry matter digestibility and nitrogen use efficiency differed across hosts, but not between populations or across temperatures. Molting accounted for 35—51% of development time. Alaskan larvae completed their fifth molt faster than Michigan larvae at 12° (11.8 vs. 17.8 d), but not at 30° (3.1 vs. 2.9 d). In both populations, molt was more temperature sensitive than growth at low temperatures (Q10 of 5.65 vs. 3.04 from 12 to 18°), but less temperature sensitive at high temperatures (Q10 of 1.60 vs. 2.06 from 18 to 30°). Survival differed across temperatures, but not between populations. Under ideal basking conditions, larvae in the field were able to elevate body temperatures °10° above ambient, but such conditions were rare in Alaska and larvae were usually near ambient temperature. Alaskan larvae were no better than Michigan larvae at selecting high radiation microsites or converting solar radiation into heat. Growth rates of Alaskan larvae were the same in the field and laboratory when fed the same foliage and exposed to the same mean daily air temperature. We incorporated P. canadensis temperature responses into a life history development model, then used a 48—yr climatic record to evaluate the fitness contributions of apparent adaptations to Alaskan summers. On a good host, at Alaskan temperatures, Alaskan P. canadensis had an estimated fitness 3.0 times greater than Michigan P. canadensis. Furthermore, the Michigan population was predicted to go extinct in 31 of 48 yr at Alaskan temperatures. Changes in growth temperature responses made the greatest contribution to enhanced fitness in Alaska, followed by increased mass of neonates, enhanced molting abilities at low temperatures, and a reduced size threshold for pupation. Analysis of fitness trade—offs suggested that extreme summers have been more important than average summers in shaping adaptive responses. Regional adaptation to climate allows P. canadensis to maintain a broader geographic distribution than would otherwise be possible, but northern distribution limits are probably still constrained by summer temperatures.