The approach to fisheries termed "balanced harvesting" (BH) calls for fishing across the widest possible range of species, stocks, and sizes in an ecosystem, in proportion to their natural productivity, so that the relative size and species composition is maintained. Such fishing is proposed to result in higher catches with less negative impact on exploited populations and ecosystems. This study examines the models and the empirical evidence put forward in support of BH. It finds that the models used unrealistic settings with regard to life history (peak of cohort biomass at small sizes), response to fishing (strong compensation of fishing mortality by reduced natural mortality), and economics (uniform high cost of fishing and same ex-vessel price for all species and sizes), and that empirical evidence of BH is scarce and questionable. It concludes that evolutionary theory, population dynamics theory, ecosystem models with realistic assumptions and settings, and widespread empirical evidence do not support the BH proposal. Rather, this body of evidence suggests that BH will not help but will hinder the policy changes needed for the rebuilding of ecosystems, healthy fish populations, and sustainable fisheries.