The quantitative accuracy of SPECT is limited by photon attenuation and scatter effect when photons interact with atoms. In this study, we developed a new attenuation correction (AC) method, CT-based mean attenuation correction (CTMAC) method, and compared it with various methods that were often used currently to assess the AC phenomenon by using the small-animal SPECT/CT data that were acquired from various physical phantoms and a rat. The physical phantoms and an SD rat, which were injected with (99m)Tc, were scanned by a parallel-hole small-animal SPECT, and then they were imaged by the 80 kVp micro-CT. Scatter was estimated and corrected by the triple-energy window (TEW) method. Absolute quantification was derived from a known activity point source scan. In the physical-phantom studies, we compared the images with original, scatter correction (SC) only, and the scatter-corrected images with AC performed by using Chang's method, CT-based attenuation correction (CTAC), CT-based iterative attenuation compensation during reconstruction (CTIACR), and the CTMAC. From the correction results, we find out that the errors of the previous six configurations are mostly quite similar. The CTMAC needs the shortest correction time while obtaining good AC results.