In previous studies, we developed linear regression models to age-grade female Aedes aegypti L. reared and maintained under controlled laboratory conditions. The models were based on temporal differences between two cuticular hydrocarbons, pentacosane (C25H52) and nonacosane (C29H60), which were extracted from Ae. aegypti legs and analyzed by gas-liquid chromatography. These initial models predicted adult female age up to 165 DD (12-15 calendar d at 28 degrees C). The age of older mosquitoes, however, could not be accurately predicted. In this study, our original regression models were tested using age data obtained from mosquitoes maintained in a field laboratory and those that were marked, released, and recaptured in northwestern Thailand. Our field data led to the development of two new regression models: one for the cool-dry season (February-March) and one for the rainy season (July-August). Both models resulted in better estimates of age than the original model and thus improved our ability to predict the age of Ae. aegypti to 15 calendar d. Females older than 15 d can be identified as such, but their exact age cannot yet be estimated. The new models will be useful for epidemiological studies and evaluating the impact of Ae. aegypti control interventions for disease prevention.