Literature on forgetting in prose learning was reviewed to establish the relative importance of five variables: (1) the definition of prose and the organizational structure of the material implied in the definition; (2) the time factor; (3) the mode of criteria measurement; (4) the similarity factor; and (5) strength of learning. Each of these was considered in turn and its relative contribution assessed. It was found that (1) all kinds of stimulus materials have been used in studying interference effects in prose learning--the most common type being constructed passages with identifiable elements that would be switched for interpolated learning and later directly tested for retention; (2) similarity seemed to be a determining variable as in paired-associate learning; (3) timing seemed to be an important factor; (4) the mode of testing influenced quantitatively the measured degree of learning and recall; and (5) strength of original and interpolated learning appeared to be a significant factor. A bibliography is included. (AW)