Quality control, as applied in normal activation analysis by the simultaneous analysis of well-characterized quality control samples, blanks and sometimes duplicates, cannot fully be applied in large sample analysis. Well characterized control samples are, e.g., not available at the size of large samples. Different approaches have to be developed to monitor and to control sources of errors in this new type of chemical analysis. Some of the measured sample parameters dealing with gamma-ray and neutron attenuation can only vary between well known values of elemental constants. These parameters can much easier be determined in large sample analysis than with samples in the milligram range, thus offering an outlook for direct verification of the quality of the related correction algorithms. Some examples are given here from experience with the kg-scale neutron activation analysis at the facilities in Delft.