The Ranschburg effect refers to the finding of impaired serial recall of items repeated on a list. One account attributes this effect to the use of a strategy where subjects avoid using as guesses stimuli that they had recalled elsewhere on the list. Support for this interpretation is reported here. The Ranschburg effect is eliminated when subjects are instructed to avoid guessing. Also, the Ranschburg effect is found in partial report only when subjects are told that the crucial item occurred elsewhere on the list.Immediate serial recall of a list of items is impaired when an item is repeated on a list. This phenomenon, known as the Ranschburg effect, was introduced into the modem memory literature by Crowder and Melton (1965). Two important qualifications are that only recall of the second occurrence of a repeated item is generally impaired and that this effect is typically found only when one or more other items intervene between the two occurrences of a repeated stimulus (Crowder, 1968a). Thus, the Ranschburg effect violates two established principles of memory-namely, that repetition improves memory and that the beneficial effects of repetition improve as a function of spacing.Two approaches have generally been taken to explain this phenomenon. The first stresses the importance of output interference in memory. This approach claims that the emission of a stimulus as a response somehow makes a second occurrence of that stimulus harder to retrieve. There are several findings supporting the role of output processes. For example, although recall of the second occurrence is impaired when subjects recall a list of items in typical forward fashion, recall of the first occurrence may be impaired when subjects recall a list in backward order (Jahnke, 1969). No Ranschburg effect is found when subjects are given memory tests that do not require emission of an item twice, such as list recognition (Wolf & Jahnke, 1968) or probed recall of single items (Jahnke, 1970). Also, Ranschburg effects may be caused by response prefixes; when subjects have to emit a stimulus before recall, they are more likely to make errors if that stimulus also occurs on the list (Crowder, 1968b).A second approach does not deny the importance of output processes in this effect but claims that this is not truly a memory phenomenon at all. Rather, the Ranschburg effect is interpreted as a result of the guessing strategies that a subject follows when he/she is unable to remember