A distinction is made between fitness‐enhancing aspects of a trait and selective force(s) that result in its optimization. Any functional explanation conceived to account for the occurrence of mobbing behaviour must explain both the benefits for the initiator and for the birds joining it. Nine teleonomic hypotheses are proposed that permit predictions about the behaviour of the prey and/or the predator in terms of mobbing or its subsequent effects on the behaviour of both parties. Predictions of different hypotheses refer to different aspects of mobbing when prey and predator are considered concurrently. Only two hypotheses lead to contrasting predictions, i. e. to “strong inference tests”. The weakness of the teleonomic method is that it does not enable us to determine when the list of possible functions of a behaviour is exhausted.