There are three hypothesized controls on food-chain length (FCL): energy supply (or "resource availability"), ecosystem size and disturbance (or "environmental variation"). In this article, the evidence for controls on FCL in freshwater ecosystems is evaluated. First, the various ways FCL can be measured are defined. Food-chain length typically is estimated as (1) connectance-based FCL - an average connectance between basal resources and top consumers, (2) functional FCL - by experimental determination of functionally significant effects of a top predator on lower trophic-level biomass patterns, and (3) realized FCL - an average connectance measure weighted by energy flow between basal consumers and the consumer occupying the maximum trophic position in the food web. Second, all evidence for relationships between the three hypothetical controls and FCL in freshwater ecosystems are evaluated. The review includes studies from streams, lakes, ponds, wetlands, phytotelmata, and experimental containers. Surprisingly, few studies of FCL in freshwaters that test the same suite of controls using the same methods are found. Equally compelling results arise from case studies based on functional, realized, and connectance-based measures of FCL. Third, 10 rules of thumb that could increase similarity of future studies, thereby facilitating synthesis across systems, are suggested. Fourth, it is discussed how FCL influences the concentration of contaminants in large-bodied animals (many of which are consumed by humans) as well as the efficacy of biocontrol applications in agriculture. Finally, there is a discussion of the potential relationships between global climate change, hydrology, and FCL in freshwaters.