Three-dimensional (3D) thermal resistance analysis provides a rapid and simple method to estimate the power generated from a waste heat recovery system with thermoelectric generators (TEGs), and facilitates an optimization of the system. Such a system comprises three parts a waste heat recovery chamber, TEG modules and a cooling system. A fin-structured duct serves as a waste heat recovery chamber, which is attached to the hot sides of the TEGs; the cold sides of the TEGs are attached to a cooling system. The waste heat recovery chamber harvests energy from exhaust heat that the TEGs convert into electricity. The estimation of generated power is an important part of the system design. Methods of Computational Fluid Dynamics (CFD) assist the analysis and improve the performance with great accuracy but great computational duration. The use of this method saves much time relative to such CFD methods. In 3D thermal resistance analysis, a node of unknown temperature is located at the centroid of each cell into which the system is divided. The relations of unknown temperatures at the cells are based on the energy conservation and the definition of thermal resistance. The temperatures of inlet waste hot gas and ambient fluid are known. With these boundary conditions, the unknown temperatures in the system are solved, enabling estimation of the power generated with TEGs. A 3D model of the system was simulated with FloTHERM; its numerical solution matched the solution of the 3D thermal resistance analysis within 6%. The power generated with the same system with TEGs (TMH400302055, Wise Life Technology, Taiwan) was measured; the experimental result is consistent with the result obtained from the 3D thermal resistance analysis; the relative deviation is approximately 10%. The power generated is affected by many variables; the positions of the TEGs, the uniformity of the internal flow of the velocity profile and the internal and external flow velocities are considered in our 3D thermal resistance analysis. According to the results, both the positions of the TEGs and the uniformity of the internal flow of the velocity profile should be taken into account to maximize the power generation. Under varied operational conditions, the power generated from the system might be more sensitive to the velocity of either the internal or external flow. Choosing an appropriate method makes increasing the power generation efficient. The relations between variables and power generation are readily revealed, even with varied parameters, yielding an optimal design of a waste heat recovery system. (C) 2016 Elsevier Ltd. All rights reserved.