We propose a quality-of-service (QoS) driven power and rate adaptation scheme for multichannel communications systems over wireless links. In particular, we use multichannel communications to model the conceptual architectures for either diversity or multiplexing systems, which play a fundamental role in physical-layer evolutions of mobile wireless networks. By integrating information theory with the concept of effective capacity, our proposed scheme aims at maximizing the multichannel-systems throughput subject to a given delay-QoS constraint. Under the framework of convex optimization, we develop the optimal adaptation algorithms. Our analyses show that when the QoS constraint becomes loose, the optimal power-control policy converges to the well-known water-filling scheme, where the Shannon (or ergodic) capacity can be achieved. On the other hand, when the QoS constraint gets stringent, the optimal policy converges to the scheme operating at a constant-rate (i.e., the zero-outage capacity), which, by using only a limited number of subchannels, approaches the Shannon capacity. This observation implies that the optimal effective capacity function decreases from the ergodic capacity to the zero-outage capacity as the QoS constraint becomes more stringent. Furthermore, unlike the single-channel communications, which have to trade off the throughput for QoS provisioning, the multichannel communications can achieve both high throughput and stringent QoS at the same.