In this paper we develop a multiagent simulation model to explore the impact of learning dynamics on the productive implementation of innovations in project networks comprised of designers and contractors. Though researchers generally agree that when firms form into networks they will gain access to new knowledge, the question of learning beyond the firm at the boundaries between firms or at the level of the project network itself remain less explored. We simulate the impact of task interdependence and relational instability on learning in project networks comprised of multiple specialist firms. We find that relational instability in project networks slows learning and task interdependence moderates the impact of increasing relational instability on network productivity rates. The findings have significant implications for interorganizational learning theory. The simulation results provide insights into appropriate firm-level and project network-level strategies for change.