The mechanistic link between metabolic rate and foraging behaviour is a crucial aspect of several energy‐based ecological theories. Despite its importance to ecology however, it remains unclear whether and how energy requirements and behavioural patterns are mechanistically connected. Here we aimed to assess how modes of behaviour, including cumulative space use, patch selection and time spent in an experimental resource‐patchy environment, are related to a forager's standard metabolic rate (SMR) and its main determinants, i.e. body mass and temperature. We tested the individual behavioural patterns and metabolic rates of a model organism, the amphipod Gammarus insensibilis, across a range of body masses and temperatures. We demonstrated quantitatively that body mass and temperature exert a major influence on foraging decisions and space use behaviour via their effects on metabolic rates and the marginal value of energy. Individual cumulative space use was found to scale allometrically with body mass and exponentially with temperature, with patch giving‐up time falling as body mass and temperature increased. In response to warmer temperatures, the specimens adaptively intensified their foraging effort and explored larger spaces to offset the elevated SMR. Our results showed that SMR explained more variation than body mass and temperature combined, and had greater predictive power for behavioural patterns. Furthermore, foraging decisions regarding patch choice and partitioning were strongly related to mass‐and‐temperature‐adjusted SMR (residual), which is a component of metabolic phenotype. Individuals with higher M–T adjusted SMR initially preferred the most profitable patch and, as time progressed, abandoned the patch earlier and explored more space than the others. These results demonstrate that foraging decisions are intimately associated with variations in standard metabolic rate, whether phenotypic or due to size and temperature combined. This, in turn, sheds light on higher‐order energy‐based ecological processes.