Facultative paedomorphosis is the ability of a salamander to either metamorphose into a terrestrial, metamorphic adult or retain a larval morphology to become a sexually mature paedomorphic adult. It has been hypothesized that density and initial body size variation within populations are instrumental in cueing metamorphosis or paedomorphosis in salamanders, yet few studies have adequately tested these hypotheses in long-term experiments. Beginning in the spring of 2004, 36 experimental ponds were used to manipulate three body size variation levels (low, medium, high) and two density levels (low, high) of Ambystoma talpoideum larvae. Larvae were individually marked using visible implant elastomers and collected every 2 weeks in order to measure snout–vent length and mass. Bi-nightly sampling was used to collect new metamorphs as they appeared. Analysis revealed significant effects of density, size variation and morph on body size of individuals during the summer. Individuals that metamorphosed during the fall and following spring were significantly larger as larvae than those becoming paedomorphic across all treatments. These results support the Best-of-a-Bad-Lot hypothesis, which proposes that the largest larvae metamorphose in order to escape unfavorable aquatic habitats. Large larvae may metamorphose to leave aquatic habitats, regardless of treatment, due to the colder climate and lower productivity found in Kentucky, which is in the northern-most part of A. talpoideum’s range. By maintaining a long-term experiment, we have provided evidence for the transition of both larvae and paedomorphs into metamorphs during fall and spring metamorphosis events. Furthermore, the production of similar morphs under different environmental conditions observed in this research suggests that the ecological mechanisms maintaining polyphenisms may be more diverse that first suspected.