Previous investigations have shown that participants are biased to respond "possible" to studied items when asked to decide whether objects could or could not exist in an object possibility test. The present study clarified and extended the concept of bias in implicit memory research in two ways. First, the authors showed that participants were biased to respond "possible" (rather than "impossible") on the object possibility test because structural processing was facilitated by prior study of possible, but not impossible, portions of objects. Second, the authors demonstrated that bias in this context was a form of, not an alternative to, implicit memory, by showing priming effects in response times when accuracy scores for studied and unstudied items were equated. The authors concluded by comparing proceduralist and memory-systems accounts of implicit memory effects and suggested that the two approaches could be seen as complementary rather than conflicting.