This systematic review was conducted to evaluate the image quality performance when implementing computed tomography data (CTAC) or magnetic resonance data for attenuation correction (MRAC) on positron emission tomography (PET) images. The CTAC and MRAC were performed on image from PET/CT and PET/MR scanners, respectively. The systematic review was done based on Preferred Reporting Items for Systematic Reviews (PRISMA). In this study, twelve articles were included from six databases. The image performance was evaluated by overall image quality, contrast, spatial resolution, detectability, standardised uptake value (SUV) and acquisition time. Data was shown as mean ± standard deviation and compared between CTAC and MRAC images to determine which attenuation correction method provides better image quality. Results found that PET-CTAC and PET-MRAC have similar image performance in overall image quality (p=0.93), detectabilty (p=0.84), SUVmean (p=0.84) and SUVmax (p=0.81). Meanwhile, PET-CTAC acquisition time is significantly faster than PET-MRAC by approximately two fold (p <0.05). There were no statistical analyses performed for image contrast, spatial resolution and contrast-noise-ratio due to the insufficient data. In conclusion, although PET/CT is faster than PET/MRI procedure, images yielded from CTAC and MRAC are equivalent to each other. Due to the variation of linear attenuation coefficient for each type of tissue, future review of image quality comparison can be done focusing on specific tissue or region such as soft tissue, bone and lungs to reflect the real impact of CTAC and MRAC on PET image.