Detailed chemical kinetics mechanisms describing low-temperature combustion often include thousands of species and reactions due to the abundance of intermediates and their complex potential energy surfaces. While these mechanisms are necessary for accurately predicting species concentrations and global combustion metrics, they are often too large for practical engine simulations that require computational fluid dynamics. As such, strategies have been developed to reduce the size of mechanisms while retaining the necessary species and reactions for high-fidelity simulations. There are three main types: skeletal mechanism generation, time-scale analysis, and isomer lumping. The first step in mechanism reduction is skeletal mechanism generation, of which directed relation graph is one of the most widely used strategies in the literature. For further reduction, time-scale analysis on the skeletal mechanism via the application of quasi-steady-state assumptions is typically employed. As an alternative, isomer lumping can be used to combine multiple isomers into one representative species. In this seminar, I will discuss the merits and disadvantages, in addition to the theory, of each strategy. I will also assess how well the final reduced mechanism derived from each strategy, or combination of different strategies, predicts global combustion properties, in comparison to the detailed mechanism. Finally, I will discuss the future of mechanism reduction.