Purpose To compare the accuracy of a single 20-second deep-inspiration breath hold (DIBH) in fluorodeoxyglucose (FDG) positron emission tomography (PET)/computed tomography (CT) to that with conventional free-breathing (FB) whole-body PET/CT for the assessment, characterization, and quantification of lung lesions in terms of the blurring effect of respiratory motion. Materials and Methods Institutional review board approval was obtained, and the requirement to obtain informed consent was waived. A preclinical study was performed in a test population of 19 patients to evaluate the feasibility and consistency of DIBH techniques compared with phase-based respiratory gating (PBRG). One hundred fifteen patients with lung lesions were then prospectively included and assessed with FB PET/CT followed by 20-second DIBH PET/CT. Maximum standardized uptake value (SUVmax), peak standardized uptake value (SUVpeak), and number and size of nodules were reported for each acquisition and then compared with findings from histopathologic examination and/or clinical-radiologic follow-up. Statistical analysis was performed with the t test, χ2 test, Pearson correlation coefficient, and receiver operating characteristic analysis. Results In the test population, data obtained with DIBH PET and PBRG PET showed close correlation (r = 0.94, P < .001 for SUVmax and r = 0.98, P < .001 for SUVpeak). In the clinical population, both SUVmax and SUVpeak were significantly increased with DIBH compared with FB (5.60 ± 4.20 vs 3.11 ± 1.80 and 2.25 ± 1.75 vs 1.71 ± 0.96, respectively; P < .001). A significantly greater number of lung lesions was detected with DIBH PET/CT compared with FB PET/CT (P < .001), with the detection of 70 additional nodules and more accurate coregistration of 84. According to the area under the receiver operating characteristic curve for SUVpeak, DIBH demonstrated a higher level of accuracy than did FB (P = .039). Conclusion The DIBH PET/CT technique is feasible in routine clinical practice and is more sensitive for quantitative measurements and lesion localization. This technique reduces the blurring effect of respiratory motion, thus improving the diagnostic accuracy for lung nodules. © RSNA, 2017.