The capacity of wireless ad hoc networks has been studied in an excellent treatise by Gupta and Kumar [1], assuming a fixed transmission rate. By contrast, in this treatise we investigate the achievable throughput improvement of rate adaptation in the context of random ad hoc networks, which have been studied in conjunction with a fixed transmission rate in [1]. Our analysis shows that rate adaptation has the potential of improving the achievable throughput compared to fixed rate transmission, since rate adaptation mitigates the effects of link quality fluctuations. However, even perfect rate control fails to change the scaling law of the per-node throughput result given in [1], regardless of the absence or presence of shadow fading. This result is confirmed in the context of specific adaptive modulation aided design examples.