PURPOSE: The objective of this phantom study was to determine whether breathing-synchronized, silicon photomultiplier (SiPM)-based PET/CT has a suitable acquisition time for routine clinical use. METHODS: Acquisitions were performed in list mode on a 4-ring SiPM-based PET/CT system. The experimental setup consisted of an external respiratory tracking device placed on a commercial dynamic thorax phantom containing a sphere filled with [F-18]-fluorodeoxyglucose. Three-dimensional sinusoidal motion was imposed on the sphere. Data were processed using frequency binning and amplitude binning (the "DMI" and "OFFLINE" methods, respectively). PET sinograms were reconstructed with a Bayesian penalized likelihood algorithm. RESULTS: Respiratory gating from a 150-sec acquisition was successful. The DMI and OFFLINE methods gave similar activity profiles but both were slightly shifted in space; the latter profile was closest to the reference acquisition. CONCLUSION: With SiPM PET/CT systems, the amplitude-based processing of breathing-synchronized data is likely to be feasible in routine clinical practice.