Abstract.-Life history should relate to gene flow (Nm) through its influence on dispersal and effective population size. Comparative studies designed to elucidate this relationship must contend with historical events that can yield misleading estimates of gene flow and statistical problems associated with inclusion of life-history traits correlated with phylogeny. We studied the relationships of life-history characters and gene flow in 15 species of darters, a monophyletic group of stream fishes. Populations of coexisting species were sampled in three geographic regions with different Pleistocene glaciation histories. Gene flow was estimated indirectly from allozymes using two methods, 6 and private alleles. Isolation-by-distance was also tested using regression of pairwise estimates of gene flow (M) on distance. Theta and private-alleles methods produced congruent estimates of Nm, except in a study region hypothesized to have been historically fragmented and then united following Pleistocene glaciation. A relatively weak association between life-history traits and Nm (based on 6) was observed when species from the historically fragmented region were included in stepwise regression analysis, because Nm was low despite life-history differences among taxa in this region. Excluding observations from this region produced stronger associations between clutch size and Nm (r 2 = 0.57), and between female size, egg size, and Nm (r 2 = 0.95). Additional analyses that corrected for female body size and phylogenetic nonindependence agreed that darters with high fecundity and small eggs exhibited high gene flow, whereas darters with small clutches and large eggs had low gene flow. The latter combination of life-history traits primarily is exhibited in species from headwater habitats where parental investment presumably confers survivorship on offspring. Reduced gene flow and genetic divergence among demes appear to be evolutionary consequences of this strategy.Key words.-Allozymes, comparative approach, Etheostoma, life history, mitochondrial DNA, Percina, population history.Received July 7, 1998. Accepted August 18, 1998 Life history should relate to gene flow (Nm) through its influence on dispersal and effective population size. A number of comparative studies have shown that interspecific lifehistory variation and gene flow are only weakly related (reviewed in Avise 1994). This probably results from the effects of historical events on genetic structure, for example, vicariant fragmentation, range expansion, and extinction and recolonization. Such events can produce patterns of genetic divergence that are similar to patterns resulting from ongoing migration (Felsenstein 1982;Templeton et al. 1995). Thus, comparing taxa with different population histories can diminish the explanatory power of life-history traits as predictors of gene flow (Loveless and Hamrick 1984;Boileau et al. 1992; Schulman and Bermingham 1995).Comparisons of unrelated and ecologically dissimilar taxa can obscure the relationship between life-history ...