The aim of this study was to assess the frequency and severity of respiration-induced curvilinear respiration artifacts (RICA) on co-registered positron emission tomography/computed tomography (PET/CT) images acquired on a combined PET/CT scanner before and after modifying the respiration protocol for CT scanning, with retrospective analysis of two groups of 100 patients each, before and after implementing a respiration protocol with breath-hold (BH) in the normal expiration position for the acquisition of the CT images. The CT data were used as attenuation map and for image co-registration. A ranking of co-registered PET/CT and PET images (including maximum intensity projection) was done by two observers in consensus using a scale from 0 to 3. Zero indicated that no RICA was visible and 1, 2, and 3 described artifact with increasing severity. A significant difference in RICA occurrence was found between the two groups ( p<0.0001). There was a 45% decrease of artifact frequency when using the normal expiration protocol and a 68% decrease of grade-2 and grade-3 artifacts ( p=0.004). The results of this study suggest that BH during the normal expiration position for CT scanning can be recommended to reduce the occurrence and the severity of RICA on PET/CT.