SYNOPSIS. Propagule size is perhaps the most widely recognized and studied maternal effect in ecology, yet its evolution is not well-understood. The large body of extant optimality theory treats parental investment solely as an ecological problem, largely from the perspective of progeny. This approach has had limited success explaining the ubiquitous variation in propagule size within and among natural populations at most temporal and spatial scales. This problem aside, an unassailable gap in propagule size theory is that it pays little heed to the fact that offspring size is a joint phenotype of two individuals- the offspring and its mother. Hence, the ecology of mothers is decidedly as important in shaping the evolution of propagule size phenotypes. There are two reasons to suspect that this gap may account for the lack of success of optimality theory to explain variation in nature. The first is that optimality models of propagule size make no allowance for, nor can they explain, widespread, multivariate correlations between maternal characters and clutch parameters, namely the positive phenotypic covariances of maternal age, size, fecundity, and per-propagule investment found in many organisms. If per-propagule investment is optimized by selection based on the expectation of offspring fitness, then why should that phenotype be a function of maternal age or size when the ecological circumstances of progeny are not changing as a function of maternal age or size? The second gap in current theory is that, like all optimization theory, it is patently non-genetic in that it is assumed that the phenotypes optimized are evolutionarily accessible. Recent maternal effects theory indicates that traits subject to maternal influence behave in unanticipated ways. Specifically, there may be time lags in response to selection, and hence, selection away from the optimum phenotype. This paper explores a suite of issues pertaining to the evolution of propagule size from the broader perspective of propagule size as a maternal effect (PSME) with a goal of widening the lens through which propagule size is viewed by evolutionary ecologists. Two themes are developed. First, I suggest that, to understand egg size variance and its implications for both maternal and offspring fitness, it is necessary to consider explicitly the ecological context in which a mother is producing eggs, not just that into which offspring will enter. I argue that some of the variables that have only been incorporated in pairwise fashion (or not at all) into studies of propagule size might account for the lack of agreement about how this important life history feature evolves. Further, I suggest that failure to consider other sources of selection on maternal phenotypes, driven by a narrow adaptationist view that has historically been taken of PSMEs, has obfuscated many interesting questions surrounding their coevolution with maternal characters. Thus, the second theme is that it is necessary to consider other explanations for why prop-agule size varies apart from those pertaining to offspring fitness per si. Based on a detailed review of the empirical literature, I conclude that the concept of an optimal propagule size is not only an insufficient construct to explain the evolution of propagule size, but that continued reliance on an optimization approach is likely to stifle development of more realistic and predictive theory for the evolution of this key life history trait. Novel theory should incorporate realities from physiology, development and genetics and should accommodate the dynamic nature of the selective environments in which propagule size evolves, all of which have been shown by empiricists to play a role in determining propagule size phenotypes. A key feature of this theory should be the explicit treatment of propagule size as a maternal effect.