Cost reduction strategies are often invoked as explanations when studies of adaptation fail to find predicted costs. This might seem discouraging, offering little opportunity for further investigation. In this paper, we demonstrate that cost reduction strategies can themselves be investigated by arguments from design. Recent work on inducible morphological defences has shown that hydrodynamical disadvantages (e.g. high drag) in fishes can be compensated for by standard metabolic rate (SMR) adjustments. Here, we theoretically investigate the possibilities and limitations for swimming cost compensation through SMR adjustment. We continue by modelling how intraspecific power curve variation affects the optimal swimming velocity between food patches. Our results show that, even though SMR modifications may compensate for hydrodynamical disadvantages, low–drag fishes will nevertheless have a marked advantage under high food abundance. The relative advantage will decrease with decreasing food levels. We also show that hydrodynamical properties of fishes can be used to predict their propensity to become foraging (or swimming) specialists. Low–drag fishes can use a broad range of swimming velocities without substantial increases in swimming cost, whereas the cost of deviating from the optimal swimming velocity increases markedly in high–drag fishes. The results have important implications for the evolution of morphological diversity in fishes.