This paper, the second of two concerned with the explicit modelling for statistical procedures for comparing members of a series of radiocarbon age determinations, considers use of a maximum likelihood radio test to cluster groups of age determinations. Three distinct case situations are defined. They represent typical problems encountered in the treatment of radiocarbon data by archaeologists and others. Use of the recommended procedures is illustrated by their paradigmatic application to experimental data, and a computer program for performing the required calculations is provided.