The main thrust of this study is to investigate methods of optimising the radiation dose–image quality relationship in computed radiography (CR) systems for chest imaging. Specifically, this study investigates the possibility of reducing the patient radiation exposure through an optimal selection of tube filtration, exposure parameters and air gap technique, in parallel with a study of the image quality, particularly low contrast detail detectability, signal-to-noise ratio (SNR) and scatter fraction (SF). The CDRAD phantom was used to assess the quality of the CR images. Tissue equivalent Polystyrene blocks were placed in the front of the phantom as scattering material with thicknesses of 5 and 15 cm to simulate an adult chest and heart/diaphragm regions, respectively. A series of exposure techniques were used, including Cu filtration with various thicknesses of Cu in the presence and absence of an air gap, whilst the exposure was kept as constant as possible throughout. The estimated patient effective dose and skin entrance dose were calculated using the NRPB-SR262 X-ray dose calculation software. The results have shown that the low contrast-detail detectability in the lung and the heart/diaphragm regions improves when using an air gap and no Cu filtration, particularly at low kilovoltage (kVp). However, there is no significant difference in low contrast-detail in the absence or presence of a 0.2 mm Cu filtration. SF values for the lung and heart regions decrease when using both, the air gap technique and a 0.2 mm Cu filtration, particularly at low kVp. SNR values for the lung and heart regions improve when using a small Cu thickness. In conclusion, this investigation has shown that the quality of chest CR images could be improved by using an air gap technique and a 0.2 mm Cu filtration at low kVp, particularly at 99 kVp.