Strength theory can account for certain memory phenomena only if strength is regarded, not as the basic unit of memory, but as some overall measure of the amount of information in store about a particular item. However not all information need be stored directly—information about recency and frequency of presentation is potentially derivable from the strength of an item rather than from recency or frequency information per se . This paper examines the evidence for and against such strength theories of recency and frequency, and reports two experiments. Strength theories of frequency are almost completely discredited. The case against strength theories of recency is less compelling—subjects may revert to using strength as the basis for recency judgments if specific temporal or serial information is lacking.