Increasingly stringent legislations call for more efficient and cleaner combustion technology as well as sustainable fuels. Chemical kinetic models are required in designing and optimizing novel engine concepts as well as selecting appropriate renewable fuels. Among the many reactions controlling fuel reactivity, OH + Fuel elementary reaction is one of the most important reactions that plays a critical role from low to high temperatures. In this thesis, OH + Fuel elementary reactions are studied for a wide spectrum of conventional and renewable fuels. The overall rate coefficients are measured in a shock tube using OH time-history profiles recorded with a UV laser diagnostic. Alkanes constitute important components of gasoline and diesel. Overall rate coefficients are measured for a series of large branched alkanes and the rate rules are derived based on the next-nearest-neighbor classification method. The strength of this method lies in the ability to predict the rate coefficients for large and/or highly-branched alkanes, where both experiments and theoretical calculations are hard to reach. Next, OH reactions with bio-derived fuels, methanol and cyclic-ketones, are studied. For OH + methanol reaction, site-specific contributions from different C-H bonds are quantified using deuterium kinetic isotopic effect, and the measured rate coefficients are found to improve the general behavior of a detailed methanol kinetic model. Reactions of cyclic ketones with OH radicals are found to exhibit similar reactivity as those of similar carbon length acyclic ketones + OH reactions. Acetaldehyde is one of the most abundant hazardous byproducts in the combustion of various fuels. Similar to methanol, OH + acetaldehyde reaction is 4 studied at the site-specific level and the importance of competing reaction channels are quantified at high temperatures. Finally, reactions of OH + cyclohexadienes and OH + trimethylbenzenes, relevant for the fate of polycyclic aromatics hydrocarbons, are investigated. A highly complex temperature dependence is observed for these molecules, a six-parameter Arrhenius expression is needed to describe the overall reactivity. The work reported in this thesis provides elementary reaction data that are highly valuable for increasing the fidelity and accuracy of predictive chemical kinetic models.