A misunderstanding of evolution via natural selection has led many plant physiologists and genetic engineers to look in the wrong direction for higher-yielding crop genotypes. Large investments in attempts to make ‘better’ plants by improving basic physiological processes are not likely to succeed because natural selection has been optimizing these for millions of years. Increases in yield from plant breeding have usually resulted from decreases, not increases, in plant fitness. Examples include reduced plant height and more vertical root growth in cereals. Plant scientists and breeders should generate hypotheses based on what evolutionary biologists call ‘group selection’, looking for attributes that increase yield in ways that decrease fitness, rather than attempting to improve upon the achievements of natural selection.