This letter presents an analytical framework for radio link level performance evaluation in a wireless network using adaptive modulation and coding (AMC) and automatic repeat request (ARQ)-based error control. Both the cases of finite and infinite buffer sizes at the radio link layer are considered when the packet arrival process is modeled by a batch Markovian arrival process (BMAP), which can capture correlation in the arrival process. Using the model, radio link level performance measures such as average delay, buffer overflow probability, packet loss rate, and average spectral efficiency can be obtained, and the impacts of channel Parameters on the performance measures can be determined. Using the queue length distributions for finite and infinite buffer cases, the buffer size can be designed such that the packet overflow probability remains below the desired level. Such a cross-layer analytical framework would be very useful for network designers.