We analyze a process-based temperature model for the length distribution and population over time of mayfly nymphs. Model parameters are estimated using a Markov Chain Monte Carlo parameter estimation method utilizing length distribution data at five different stream sites. Two different models (a standard exponential model and a modified Weibull model) of mayfly mortality are evaluated, where in both cases mayfly length growth is a function of stream temperature. Based on model-data comparisons to the modeled length distribution and the Bayesian Information Criterion, we found that approaches that length distribution data can reliably estimate 2–3 model parameters. Future model development could include additional factors include such as upstream environmental factors, abiotic conditions, interspecific competition, predation, or stream salinity. Outputs of this model could be applied to predict mayfly emergence across a geographic domain or to forecast mayfly population responses to climate change.