ABSTRACT
In theory, the subject of multivariate reference regions (MRRs) for biochemical analytes follows naturally after univariate reference limits. In practice, however, the MRR remains a statistical curiosity to the great majority of clinical chemists and clinicians and has little or no use in present-day clinical practice. A rare exception like the trivariate region of thyroid tests developed by Kagedal et al. (1978, 1982) seems only to prove the rule.
