A major challenge in current research into aging using model organisms is to establish whether different treatments resulting in slowed aging involve common or distinct mechanisms. Such treatments include gene mutation, dietary restriction (DR), and manipulation of reproduction, gonadal signals and temperature. The principal method used to determine whether these treatments act through common mechanisms is to compare the magnitude of the effect on aging of each treatment separately with that when two are applied simultaneously. In this discussion we identify five types of methodological shortcomings that have marred such studies. These are (1) submaximal lifespan-extension by individual treatments, e.g. as a result of the use of hypomorphic rather than null alleles; (2) effects of a single treatment on survival through more than one mechanism, e.g. pleiotropic effects of lifespan mutants; (3) the difficulty of interpreting the magnitude of increases in lifespan in double treatments, and failure to measure and model age-specific mortality rates; (4) the non-specific effects of life extension suppressors; and (5) the possible occurrence of artefactual mutant interactions. When considered in the light of these problems, the conclusions of a number of recent lifespan interaction studies appear questionable. We suggest six rules for avoiding the pitfalls that can beset interaction studies.