Developmental times, survival, and adult dry weights for two clones of the pea aphid,Acyrthosiphon pisum(Harris), were determined at constant temperatures from 5 to 28°C1. These data were used to test the hypotheses that genetic variation in the developmental threshold or the effect of temperature on the nutritional value of food could account for nonlinearity in the relationship between developmental rate and temperature at low temperatures. Neither of these hypotheses was supported by the data and the relationship probably is inherently nonlinear. The consequences of this nonlinearity were investigated in relation to estimating developmental thresholds. Developmental thresholds based on the traditional linear method of estimation are not adequate when they are used to quantify a life history trait, but probably are adequate in pest management. Alternative methods for estimating a developmental threshold and degree-day requirement were developed based on the parameters of a nonlinear equation. The approach was validated by comparing methods for 21 diverse species of insects.