Inclusion concentrations are typically <1ppm, but many wrought products are sensitive to their presence. Their sparseness makes reliable quantification difficult. The PoDFA test is over fifty years old, but still regarded as the “gold standard”, albeit time-consuming. Its more recent derivative, the Prefil test, has attracted criticism over its reproducibility. This paper quantifies the measurement error in PoDFA analysis, raising questions over its applicability to clean metal. The Prefil test should have greater sensitivity, but two serious limitations are discussed. The filter permeability variability can dominate flow characteristics in clean metal: however post-facto curve analysis can compensate for this. Some alloys also attack the filter binder: another source of measurement error. The combination of Prefil and curve analysis (with the filtration residue available for subsequent PoDFA analysis) can provide a reliable, rapid response quality control tool which can be seen as a cost-effective alternative to LiMCA.