Numerous mathematical models have been proposed for prediction of baroreflex regulation of heart rate. Most models have been designed to provide qualitative predictions of the phenomena, though some recent models have been developed to predict observed data. In this study we show how sensitivity and correlation analysis can be used for model reduction and for obtaining a set of identifiable parameters that can be estimated reliably given a model and an associated set of data. We show that the model developed by Bugenhagen et al. to predict heart rate dynamics in the Dahl SS rat can be simplified significantly, without loss of its ability to predict measured data.