Building performance simulation has the potential to quantitatively evaluate design alternatives and various energy conservation measures for retrofit projects. However before design strategies can be evaluated, accurate modeling of existing conditions is crucial. This paper extends current model calibration practice by presenting a probabilistic method for estimating uncertain parameters in HVAC systems for whole building energy modeling. Using Markov Chain Monte Carlo (MCMC) methods, probabilistic estimates of the parameters in two HVAC models were generated for use in EnergyPlus. Demonstrated through a case study, the proposed methodology provides predictions that more accurately match observed data than base case models that are developed using default values, typical assumptions and rules of thumb.