Environmental sex determination may yield biased sex ratios, as is the case in many reptilian species with temperature-dependent sex determination (TSD). Determin- ing TSD thermal ecology is important to understanding sex ratio evolution in those species, as TSD systems may pose a challenge to Fisherian models that predict the production of 1:1 sex ratios through frequency-dependent selection. Unfortunately, the relationship be- tween fluctuating temperatures and sex in nature remains poorly understood. Here I present a novel statistical model of TSD in the endangered Amazonian freshwater turtle, Podocnemis expansa, designed to account for heterogeneous daily fluctuations of natural nest temper- atures by comparing laboratory and field data on sex ratios, developmental stages, cu- mulative temperature units (CTUs), and time spent below the survival threshold (288C). Experiments encompassed constant, shift-once, and shift-twice temperatures in the labo- ratory, plus depth and shading treatments in the field. A constant temperature of 32.68C produced 1:1 males : females, the highest pivotal temperature ever reported for a turtle species. The thermosensitive period for sex determination (TSP) roughly occurs during developmental stages 17-24 as determined from shift experiments. The CTUs during the TSP plus the number of hours below 288C prior to the TSP significantly explained sex ratio production in the combined laboratory and field data set (positive and negative effects, respectively). However, the relationship for field nests only was not significant. Higher temperatures had a disproportionately larger effect on female production in the combined field and laboratory data set. Depth and shading had drastic effects on nest temperature, embryonic development, and sex ratio. Seasonal climate changes can override the effect of nest depth via differences in oviposition time. Two of the shaded nests experienced the coldest temperatures of all field nests but yielded the highest female proportion, according to a TSDII pattern (when females are produced at low and high temperatures and males are produced at intermediate temperatures). Thus Podocnemis expansa may be a TSDII species, but this intrinsic system seems to be realized as TSDIa in nature (males produced at low temperatures and females produced at high temperatures), because of the general lethality of cold temperatures which precludes the production of low-temperature females.