Cot analysis (DNA reassociation kinetics) has long been used to explore genome structure in individual species, estimate genome similarity among organisms, and evaluate diversity in ecological samples, yet the algorithms and computational tools designed for analyzing Cot data are outdated, difficult to use, and prone to error. We report a new nonlinear regression procedure for analysis of Cot data and describe our algorithms in detail. Our procedure is implemented as CotQuest, a suite of scripts designed for use with the statistics package SAS. Unlike previous programs, CotQuest does not require users to input guesses as to the final values of parameters; rather, it employs a novel algorithm to step through a sequence of progressively more complex models, with the results from a given analysis being used to generate starting values for the next model. Moreover, CotQuest returns a statistical comparison of potential models and provides a variety of model assessment and selection diagnostics to help users in model selection. In situations where two models possess similar goodness-of-fit assessments, visual analysis of the Cot curves and comparison of CotQuest-generated graphs and statistics reflecting the normality and homoscedasticity of residuals can be employed to make educated choices between models.