Opportunistic relay networks, much like other forms of opportunism, have simpler coding/decoding, as well as improved throughput compared with open-loop systems. However, optimal opportunistic access in relay networks requires the comparison of the instantaneous capacities of sub-networks, which unfortunately are often unknown in closed form. This is much more diffcult than, e.g., uplink or downlink diversity, where it suffices to compare link gains or simple functions thereof. We make advances in understanding the performance of a wide class of opportunistic relay networks by developing a new formula for the opportunistic diversity multiplexing tradeoff.