Atkinson cycle realized with a late intake valve closing (LIVC) and Miller cycle achieved with an early intake valve closing (EIVC) have been recognized as effective approaches for improving the gasoline engine fuel economy. In both Atkinson and Miller cycles, the engine can be designed with a higher geometric compression ratio for increasing the expansion work and the effective compression ratio is governed by the intake valve close (IVC) timing for the knock control. Duration of the intake event and IVC timing affect not only the pumping loss during the gas exchange, but also have strong influences on the friction torques of the intake cams and the turbulence intensities for the in-cylinder charge motion. The latter governs duration of combustion and EGR tolerance, both of which have impacts on the engine thermal efficiency. In order to better understand how Atkinson cycle and Miller cycle influence the engine fuel consumptions at different engine speeds and loads, an experimental investigation was conducted to compare influences of EIVC and LIVC on the fuel consumption of a 1.5L turbo-charged gasoline direction injection (TGDI) engine. The engine was tested with three different intake cams, covering durations of the intake event IDR = 251 degCA (base engine), 196 degCA (Miller engine), and 274 degCA (Atkinson engine). Compression ratios are 11,5:1 for the Atkinson engine and Miller engine and 9.5:1 for the base engine, achieved with piston modifications. The results of this investigation will be reported in three papers focusing respectively on characteristics of the engine friction, in-cylinder charge motions for different intake events, and combustion and fuel economy without and with EGR for the naturally aspirated mode and boost mode. The present paper is Part III of this study. Experimental results of the base engine, the Atkinson engine and the Miller engine will be reported on the fuel consumption, combustion duration, and EGR tolerance for loads/speeds on the entire engine map. The minimum brake effective fuel consumptions (BSFC) for the three engines fueled with RON93 gasoline were found to be 240 g/kWh for the base engine, 230 g/kWh for the Miller engine and 227 g/kWh for the Atkinson engine, respectively. For most engine loads, combustion durations for the Atkinson engine were about 10 degCA shorter than those of the Miller engine. The 235-g/kWh BSFC area for the Atkinson engine was much greater than that of the Miller engine, and the EGR tolerance for the Atkinson engine was about twice that of the Miller engine. Because the base engine was not designed with an EGR system, its EGR tolerance could not be compared.