Today, the potential of HCCI engines, i.e., their high efficiency with low NO and particulate emissions, is very well known. Besides this potential, the problems that are related to HCCI engines, particularly the control, are also known issues. In order to be able to develop and assess the control strategies for HCCI engines, one needs models for the control development. In addition to mean value models crank-angle resolved control-oriented computer codes have also been developed lately (e.g., Boost RT). In these codes, complex 1D gas dynamics and complex combustion models are omitted, while the in-cylinder calculation is crank-angle resolved. Simple combustion models are variations of Vibe and the combustion defined by a table. Since the HCCI combustion is controlled by the state of the gas in the cylinder, parameters of Vibe functions depend on the factors that define this state. In this numerical study, the influence of several state factors on the ROHR curve of an HCCI engine is analyzed. Also, a possibility and a methodology of using state-dependent Vibe parameters in the simulation of combustion of HCCI engines are shown. The analysis is performed on the basis of combustion data calculated by the six-zone simulation model which uses chemical kinetics and which has been tested earlier. The calculated ROHR curves are parameterized with Vibe parameters (single or double). The analysis of the dependence of Vibe parameters on influencing factors produced a list of key factors required to define the entire operating map. This analysis also showed the main trends of these influences (linear, non-linear) which helped in the process of determining design points for the definition of regression functions. From the results of the six-zone simulation model in each design point, parameters of a double Vibe function were defined, and a regression function for each parameter was found. The obtained Vibe parameter functions are tested by comparing results from the calculation that uses chemical kinetics with simple combustion model results. In addition, the limitations of the model with Vibe parameters defined in such a way were tested by running simulations in several different operating modes (e.g., different engine speed and different intake valve closure timings).