Differential recall between compared groups is discussed most often in the context of case-control studies. Cases, compared to controls, are suspected of providing a more complete report of their true exposure to an hypothesized risk factor, thereby biasing upwards the estimate of its effect. The present paper describes how differential recall can arise with any observational design in epidemiology; with any class of study variable, not only exposures; and may inflate or deflate the true value of the estimate of effect size. We list a variety of study designs and questionnaire tactics that aim to remedy these problems. The scope and magnitude of the bias created by differential recall and the efficacy of proposed remedies require further study.