Archaeologists, along with other Quaternary researchers, seldom rely upon a single radiocarbon determination to provide an estimate of the age of the phenomenon which is the object of their study. There is an evident need for an explicitly formulated procedure for comparing sets of radiocarbon determinations from the same and from adjacent strata or sites, and for combining these where statistical and archaeological criteria indicate that this combination is warranted. The present contribution provides explicit modelling for a series of recommended procedures, a critique of previous methods, and paradigms for application of the recommended procedures.