Objective To explore the feasibility of mean temporal phase images calculated from perfusion CT datasets by using CT perfusion (CTP) of liver on the third-generation dual-source CT. Methods Twenty-two consecutive patients with suspected hepatocellular carcinoma were enrolled, we retrospectively compared objective and subjective image quality, leson detectability, and radiation dose between mean temporal arterial (mTA) and mean temporal portal venous (mTPV) images which calculated from perfusion CT datasets with conventional enhanced arterial and portal venous datasets. Results ① Image quality: compared with the conventional enhancement image, the standard deviation (SD) values of CTP images on liver (arterial phase), portal vein (arterial phase), and liver (portal vein phase) were lower (P<0.05); the signal-to-noise ratio (SNR) values of CTP images on aorta (arterial phase), portal vein (arterial phase), aorta (portal vein phase), and portal vein (portal vein phase) were all higher (P<0.05), the contrast-to-noise ratio (CNR) value of CTP images on aorta (arterial phase) was higher (P<0.05). ② The subjective image quality: the subjective image quality scores of CTP images (mTA and mTPV images) were higher when compared to responding conventional enhanced arterial and portal venous datasets (P<0.05). ③ The diagnostic efficiency: the CTP images and conventional enhancement images showed all the lesions, but the diagnostic efficiency images of CTP images was better than the conventional enhancement images, both on lesions of blood supply and lack of blood supply (P<0.05). Conclusions The image quality of mTA and mTPV datasets calculated from CTP datasets are non-inferior when compared to conventional enhanced arterial and portal venous acquisitions in patients with suspected hepatic lesions. Thus, CTP could be used as a stand-alone imaging technique without additionally performed conventional arterial and portal venous CT acquisitions.