The present study presents an algorithm developed for the adaptive run-time reduction of chemical mechanisms during HCCI chemical kinetics simulation. The algorithm is incorporated in the simulation and is used to extract a subset of reactions and species from a chemical mechanism at each simulation step during run-time. According to the algorithm, only the species and reactions included in this subset are introduced into the solver, which is used to solve the system of differential equations describing the species evolution. The criteria employed in the algorithm are based on the computational effect of each reaction and specifically on its ability to produce computationally detectable changes in the participating species mass fractions. The effectiveness of the method is demonstrated by incorporating the algorithm into single-zone models simulating HCCI combustion. Chemical mechanisms of varying degree of complexity are used in the model, both reduced and detailed ones, and the results obtained with and without the use of the algorithm are compared. A parametric investigation is also conducted by varying load, EGR rate and the initial concentration of the additive NO in selected fuels, to test the ability of the algorithm to adapt to changing inlet conditions. These investigations provide an evaluation of the algorithm effectiveness in reducing the simulation time without making significant sacrifices in the species evolution predictions. The implementation of the algorithm resulted in a computation time reduction by 40-80%, without introducing noticeable errors. (C) 2014 Elsevier Ltd. All rights reserved.