The lack of data on temperature-dependent development of instars ofLiriomyza sativaeBlanchard is an impediment to research on biological control of this pest because parasitoids will likely interact differently with each instar. In this study, development times of eggs and instars ofL. sativaewere determined at constant temperatures of 20, 25, 30, and 35 ± 1°C onPhaseolus lunatusL. ‘Henderson’. Development rates of eggs and larvae increased linearly with temperature up to 35°C (r2≥ 0.95). More than 91% of the variation in development rate of first and second instars was explained by temperature, whereas temperature explained only 42% of the variation in third-instar development rate. A part of the variation in development rate of third instars may be explained by the fact that the designated end point of this stadium (emergence of third instars from leaves) occurred only during the photophase. Overall, 99.9% of larvae emerged from leaves after 95 degree-days (DD) (threshold temperature, 10°C). When the constant temperature data were used to parameterize a stochastic phenology model forL. sativae, the model predicted that 50% of the population would advance beyond the egg and first-, second-, and third-instar stages by 39.2, 67.0, 85.5, and 101.2 DD, respectively. The model predicted peak proportions of the population to be in first, second, and third stadia at 52.7, 76.1, and 93.2 DD. Pooled data from two fluctuating temperature experiments resulted in very similar estimates of timing of peak stage proportions (51.3, 74.4, and 94.1 DD, respectively). Degree-day values at peak stage proportions were used successfully to determine the temperatures required to advance members of a cohort to a particular stadium at a given time, thereby facilitating experiments examining interactions of parasitoids with each larval instar.