I develop a model of contracting under reciprocal altruism accounting for some evidence which is paradoxical from the point of view of neoclassical models with selfish actors. My model predicts the crowding-out effect observed in the Trust Game with the possibility of a fine; for the Control Game...