Data from single case designs (SCDs) have traditionally been analyzed by visual inspection rather than statistical models. As a consequence, effect sizes have been of little interest. Lately, some effect-size estimators have been proposed, but most are either (i) nonparametric, and/or (ii) based on an analogy incompatible with effect sizes from group-based studies. As part of a research program conducted by Shadish, Hedges, and Rindskopf, this paper reports a beginning step to fill the gap between effect sizes in SCDs and in group-based designs. When, as is typically the case, an SCD involves multiple participants, it is possible to use variation among participants as a denominator in an effect size measure, as is done in between-group research. A simple estimator of this type can easily be implemented using WinBUGS, a Bayesian program that has extreme flexibility.