This paper is a first step towards a better understanding of the impact of learning strategies on the fate of organizations facing environmental discontinuities. We build an agent-based model to study formally the short-run and long-run consequences of exploitation and exploration. In the model, organizations have to be individually and ecologically strong enough to survive. Previous research shows that exploitation is efficient in the short-run whereas exploration is more risky but can help organizations adapt to environmental jolts. This trade-off is here analyzed with three kinds of organizations: pure exploiters, pure explorers, and balanced learners. Our preliminary results seem to indicate that no learning strategy is best under any circumstance. We test several environmental situations, where discontinuities vary in severity and in frequency.