We compared the pre-injection transmission scan with the simultaneous emission/transmission scan (SET) using a body phantom in positron emission tomography (PET) to study the factors affecting emission (EMIS) images with different transmission (TRAN) data. The results showed that total count, region, scan method, and EMIS component influenced the bias and noise statistics. This influence was further passed on to the EMIS data and affected the reconstructed image. However, the result of fixing EMIS scan time at 10 minutes was that the noise statistics of EMIS had a greater effect on image quality than TRAN. The optimal scan counts of SET in this study were 20 Mcounts, with a scan time of about 10 minutes (85.5 MBq, (68)Ge-(68)Ga radioactivity). In clinical use, optimal collection count differs, as radiation TRAN quantity depends on the size of the subject, even with fixed radioactivity of (68)Ge-(68)Ga.