The selection of a nonwoven air filter structure that is optimal for specific conditions is not simple and is affected by many factors. Traditionally, the selection process is based on experience and professional instinct. However, this approach to decision-making lacks an objective basis for making comparisons between alternative structures. In order to optimize the effectiveness of the filter performance, the decision-making process must be placed on a rational and objective basis. A general approach to a multicriteria optimization is discussed. Linear and nonlinear models are analyzed as well as compound criteria obtained by various combinations of simple criteria. An essential component of the optimization model is the criterion function, which is a mathematical expression of the optimization criterion deduced from the goals. Each criterion represents a filter characteristic, such as filter efficiency, pressure drop, flow rate, dust-holding capacity and, in some cases, economic indices. The formulation of the optimization problem involves transforming filter performance into an equivalent mathematical model. A set of equations was developed to describe initial efficiency and pressure drop as well as efficiency and pressure drop for dust loaded nonwoven media at intermediate Reynolds numbers. Using a simple mathematical method, the optimal structure of a nonwoven filter for a given set of conditions was obtained. A compound criterion can also be used as a quality factor in order to make comparisons between filters of different structures