We propose that the concept of cognitive effort in memory is both useful and important. Cognitive effort is defined as the engaged proportion of limited-capacity central processing. It·was hypothesized that this variable might have important memorial consequences and might also be a potential confounding factor in levels-of-processing paradigms. The first experiment tested this possibility using two types of incidental-learning tasks factorially combined with two degrees of effort. It was found that high effort led to better recall than low effort, but that level-of-processing effects were nonsignificant. A second experiment clearly demonstrated the feasibility of using performance on a secondary task as an independent criterion for measuring effort, and two further experiments ruled out alternative accounts of effort effects. A reliable levels-of-processing effect was obtained in the fourth experiment in which the incidental-learning tasks were blocked. Implications and possible future applications of the cognitive effort concept are discussed.