Evolutionary transitions between sex-determining mechanisms (SDMs) are an enigma. Among vertebrates, individual sex (male or female) is primarily determined by either genes (genotypic sex determination, GSD) or embryonic incubation temperature (temperature-dependent sex determination, TSD), and these mechanisms have undergone repeated evolutionary transitions. Despite this evolutionary lability, transitions from GSD (i.e. from male heterogamety, XX/XY, or female heterogamety, ZZ/ZW) to TSD are an evolutionary conundrum, as they appear to require crossing a fitness valley arising from the production of genotypes with reduced viability owing to being homogametic for degenerated sex chromosomes (YY or WW individuals). Moreover, it is unclear whether alternative (e.g. mixed) forms of sex determination can persist across evolutionary time. It has previously been suggested that transitions would be easy if temperature-dependent sex reversal (e.g. XX male or XY female) was asymmetrical, occurring only in the homogametic sex. However, only recently has a mechanistic model of sex determination emerged that may allow such asymmetrical sex reversal. We demonstrate that selection for TSD in a realistic sex-determining system can readily drive evolutionary transitions from GSD to TSD that do not require the production of YY or WW individuals. In XX/XY systems, sex reversal (female to male) occurs in a portion of the XX individuals only, leading to the loss of the Y allele (or chromosome) from the population as XX individuals mate with each other. The outcome is a population of XX individuals whose sex is determined by incubation temperature (TSD). Moreover, our model reveals a novel evolutionarily stable state representing a mixedmechanism system that has not been revealed by previous approaches. This study solves two long-standing puzzles of the evolution of sex-determining mechanisms by illuminating the evolutionary pathways and endpoints.