Group-wise spatial normalization of fiber orientation distribution functions (fODF) is an important step in fixel-based analysis. This work compared the accuracy in matching fODFs using fODF-based and tensor-based registration. It demonstrated superior fODF matching with fODF-based registration, as expected, even in conditions that are optimal for tensor but not fODF reconstruction. Nevertheless, it was shown that tensor-based registration has the ability to spatially match fODF features rather well, though less accurately than fODF-based registration. Finally, this work demonstrated that fODF-based transformations resulted in worse matching of structural information than tensor-based transformations.