For Pt. I, see Proc. Eur. Fed. for Medical Informatics (1984). The procedures for clinical inference of a multivariate algorithm for diagnostic information systems are presented. This complements the description given in Pt. I of procedures for data acquisition and storage of the proposed algorithm. Clinical inference is viewed as a twofold process of (1) acquiring and sampling information, and organizing it in a fashion conducive to subsequent analysis; and (2) deciding about the normality or abnormality of functional health domains. Normality is defined as being within the limits of a population criterion interval. The mathematical explication of clinical inference as used in the algorithm is emphasized.