The metabolism of several animal groups declines with depth even after adjustments for size and temperature. The ''visual-interactions hypothesis'' explains this trend as the result of declining light levels that reduce the distances over which predators and prey interact. This reduction relaxes the selective pressure for locomotory capacity, and reductions in metabolic rate follow. The decline in visual interactions and metabolism is most pronounced among pelagic species, as this environment affords no refuge from predators. The visual-interactions hypothesis thus predicts less depth-related variation among benthic species. However, it has been tested primarily with pelagic animal data. Summarizing many scattered studies and data sets to further test the hypothesis, here we analyze the data for benthic and benthopelagic fishes. Oxygen consumption rates declined significantly with depth in benthopelagic and, to a lesser extent, benthic species. Trends in muscle metabolic enzyme activities generally corroborated these patterns. The anaerobic capacity of the white muscle indicated the greatest decline in pelagic and the smallest reduction in benthic species, as expected. Similar trends were not found in aerobic capacity, but this result may reflect a paucity of enzyme data for benthic species. Most of the studied fishes live off of California, where the presence of an oxygen-minimum zone may influence some of the patterns observed. This preliminary analysis of data clearly illustrates that temperature and body mass cannot explain the variability evident in metabolism. Rather, some covariate of habitat depth acts to influence metabolism in benthic and benthopelagic fishes. The general trends are explained by the visual-interactions hypothesis, but considerably more data are required to account for the variation in metabolism and lifestyle that is apparent. In particular, regional comparisons are needed to separate the influences of environmental factors, such as oxygen, which covary with depth. Rate of metabolism is of fundamental importance to ecology because it is the process of energy assimilation, transformation, and allocation, and as such, it can be used to construct models of the flow of energy and materials in an ecosystem (Smith 1992; Smith et al. 2001). Therefore, there have been many efforts to describe patterns of metabolism and their mechanistic underpinnings (Childress 1995; Gillooly et al. 2001; Clarke 2004). Many factors are known to have important effects on animal metabolic rates. Among the best described are temperature and body mass. Temperature affects metabolism, presumably through kinetic effects on reaction rates, although the precise link is poorly understood. Broadly speaking, higher tempera-tures lead to higher metabolic rates within a species (Clarke and Johnston 1999; Clarke 2004). Body size also has dramatic effects on metabolism. The classic ''mouse-to-elephant'' curve relates the variation in mammalian metabolic rates to body sizes with a power law. Although cause and generality of the relationship is actively debated (Childress and Somero 1990; Suarez et al. 2004; Glazier 2005), smaller animals generally have higher mass-specific metabolic rates than larger ones. While some have argued that the metabolic rates of everything from microbes to blue whales can be predicted based solely on size and temperature (Gillooly et al. 2001; Brown et al. 2004), more recent analyses demonstrate a wide variation in both the slopes and elevations of the scaling relationships after temperature adjustment (Glazier 2005; Seibel 2007; Seibel and Drazen in press). Metabolic rates within some groups of deep-sea organisms are much lower than in their shallow-water counterparts even after mass and temperature effects have been taken into account. Studies have shown that pelagic fishes (Torres et al. 1979), crustaceans (Childress 1975; Childress et al. 1990a), and cephalopods (Seibel et al. 1997) exhibit rapid declines in metabolic rates with depth, and the trend cannot be explained by correcting for temperature or animal mass.