In this study, a beta-type 300-W Stirling engine is developed and tested, and a non-ideal adiabatic model is built and applied to predict performance of the engine. Engine torque, engine speed and shaft power output are measured under various operating conditions. The experiments are conducted for two different working gases (air and helium) and at various charged pressures and heating temperatures. Effects of regenerator wire mesh on the shaft power output are also examined. Results show that the shaft power output of the engine is much higher using helium as the working fluid than using air. Furthermore, as the charged pressure and the heating temperature are set at 8 bars and 850 C and a No. 120 wire mesh is used in the regenerator, the shaft power of the engine can reach 390 W at 1400 rpm with 1.21-kW input heat transfer rate (32.2% thermal efficiency). The experimental data are compared with the numerical predictions to verify the theoretical model. It is found that the experimental data of the shaft power output closely agree with the numerical predictions. This implies that the theoretical model is valid and helpful in the engine design. (C) 2013 Elsevier Ltd. All rights reserved.