In this paper we test the degree of tranfer of learning between two similar, n-player coordination games with Pareto-ranked equilibria with fixed cohorts. Our results show that achieve the payoff-dominants equilibrium in one game are then able to transfer this mutual best response outcome to the new game. However, cohorts which do not converge to the best outcome in the first game are not able to recover from coordination failure in the new game. Furthermore, the effects of tranfer on collective coordination change depending on the sequence with which the two games are played.