Incubation has long been proposed as a mechanism in creative problem solving (Wallas, 1926). A new trial-by-trial method for observing incubation effects was used to compare the forgetting fixation hypothesis with the conscious work hypothesis. Two experiments examined the effects of incubation on initially unsolved Remote Associates Test (RAT) problems. Following exposure to misleading clues designed to induce initial fixation on RAT problems, versus no clues, participants were retested on problems either immediately after their first attempt (no-incubation), or after a 40-second incubation period. Resolution of initially unsolved RAT problems (fixated versus non-fixated) was examined as a function of complete interruption (Experiment 1) or partial distraction (Experiment 2). An incubation effect, that is, better resolution of initially unsolved problems retested after a delay rather than retesting immediately, was seen only in Experiment 1, in which unsolved problems were completely removed from sight. Furthermore, an incubation effect was found only for initially fixated problems, and not for problems that were not accompanied by misleading clues. The results are consistent with the forgetting fixation hypothesis (Smith & Blankenship, 1989), which states that putting unsolved problems completely out of mind allows initial fixation to dissipate, and the results indicate that the opportunity for some conscious work during incubation periods may not be optimal for resolving fixation. (Contains 3 tables, 1 figure and 1 footnote.)