A new method of analyzing the thermal performance of parabolic trough collectors (PTCs) for solar thermal applications is established using similarity principle and dimensional analysis, through which different types of PTCs can be studied via a single scaled physical model. Six dimensionless numbers for the PTC are drived and are used to build a scaled PTC model. A coupled approach combining Monte-Carlo ray-tracing method with finite-element is developed to analyze the performance of PTCs. Experimental data from the literature are employed to calibrate the numerical model. Compared with the results of four typical cases of non-scaled experimental data obtained from the literature, differences in average efficiencies of the scaled-down models are within 0.75%, thereby validating the scaled model and similarity method for analyzing PTCs with vastly different length scales. Effects of direct normal irradiance (DNI) and temperature difference between the receiver fluid and ambient air Delta T-ab on the efficiency of PTCs are further analyzed using the scaled model. The simulation results indicate that the collector efficiency increases with the augmentation of the DNI, whereas it decreases with the increase in Delta T-ab. The similarity analysis method provides a new perspective for solar-thermal research by demonstrating the possibility of performing experiments on PTC on a much-reduced length scales. (C) 2017 Elsevier Ltd. All rights reserved.