The cost of the immune response is variable and may provide a sufficient selective pressure to produce adaptations that minimise those costs under high infection prevalence. Here, using invertebrates as a model, I suggest two possible mechanisms that maintain variation in responses that covary with costs. First I propose that infection prevalence should balance costs through the selection of optimal patterns of “specific” and “non-specific” immune pathways concomitantly expressed in the immune response. Second, I propose that life history adjustments (e.g. earlier reproduction in response to infection) could have been selected to minimise the cost of successful immune responses and consequently may result in the maintenance of costly immunity.