The flavour sector of the Standard Model with its large number of parameters is still rather badly understood. A plethora of models has been developed in the past to shed some light on the origin and structure of flavour. Nevertheless, the situation is still not very satisfactory and no vastly preferred approach has been firmly established. Instead of reviewing various model attempts we want to discuss a class of predictions which many of the models have in common: sum rules. We put an emphasis on the lepton sector, but we will also briefly mention sum rules in the quark sector.