Purpose As an alternative to technetium-99m-macroaggregated albumin (99m Tc-MAA), a scout dose of holmium-166 (166 Ho) microspheres can be used prior to 166 Ho-radioembolization. The use of identical particles for pre-treatment and treatment procedures may improve the predictive value of pre-treatment analysis of distribution. The aim of this study was to analyze the agreement between 166 Ho-scout and 166 Ho-therapeutic dose in comparison with the agreement between 99m Tc-MAA and 166 Ho-therapeutic dose. Methods Two separate scout dose procedures were performed (99m Tc-MAA and 166 Ho-scout) before treatment in 53 patients. First, qualitative assessment was performed by two blinded nuclear medicine physicians who visually rated the agreement between the 99m Tc-MAA, 166 Ho-scout, and 166 Ho-therapeutic dose SPECT-scans (i.e., all performed in the same patient) on a 5-point scale. Second, agreement was measured quantitatively by delineating lesions and normal liver on FDG-PET/CT. These volumes of interest (VOIs) were co-registered to the SPECT/CT images. The predicted absorbed doses (based on 99m Tc-MAA and 166 Ho-scout) were compared with the actual absorbed dose on post-treatment SPECT. Results A total of 23 procedures (71 lesions, 22 patients) were included for analysis. In the qualitative analysis, 166 Ho-scout was superior with a median score of 4 vs. 2.5 for 99m Tc-MAA (p < 0.001). The quantitative analysis showed significantly narrower 95%-limits of agreement for 166 Ho-scout in comparison with 99m Tc-MAA when evaluating lesion absorbed dose (− 90.3 and 105.3 Gy vs. − 164.1 and 197.0 Gy, respectively). Evaluation of normal liver absorbed dose did not show difference in agreement between both scout doses and 166 Ho-therapeutic dose (− 2.9 and 5.5 Gy vs − 3.6 and 4.1 Gy for 99m Tc-MAA and 166 Ho-scout, respectively). Conclusions In this study, 166 Ho-scout was shown to have a superior predictive value for intrahepatic distribution in comparison with 99m Tc-MAA.