We study the energy-limited outage probability of the block space-time coding (STC)-based systems utilizing automatic repeat request (ARQ) feedback and adaptive power allocation. Taking the ARQ feedback costs into account, we derive closed-form solutions for the energy-limited optimal power allocation and investigate the diversity gain of different STC-ARQ schemes. Also, sufficient conditions are derived for the usefulness of ARQ, in terms of energy-limited outage probability. The results show that, for a large range of feedback costs, the energy efficiency is substantially improved by the combination of ARQ and STC techniques, if optimal power allocation is utilized. I. INTRODUCTION Automatic repeat request (ARQ) techniques are commonly used in wireless networks to combat the loss of data packets due to channel fading. In ARQ schemes, if the receiver fails in decoding the data correctly, it asks for a retransmission. Then, depending on the data retransmission approach, different types of ARQ are defined in the literature [1]-[8]. Type I is the simplest version of ARQ, where both the error-detecting and the forward error correction information are added to each message and the receiver disregards the previous messages, if received in error. In Type II (resp. Type III), a new (resp. the same) data is sent in the retransmission rounds and, in each round, the receiver combines all signals received up to the end of that round. With multiple antennas, an alternative for Types I-III ARQ is to use the space-time coding (STC) techniques in an ARQbased fashion. With a STC-based ARQ approach [9]-[14], a permuted version of the initial sub-codeword is sent in the retransmissions and, in each round, the receiver combines all received permutations of the initial signal for message decoding. Thus, the STC-based ARQ is an intermediate type of ARQ with (almost) the same complexity as in Type III and data transmission efficiency comparable with Type II ARQ [9]-[14]. Moreover, the implementation of STC-ARQ protocols is of interest when we remember that the STCs are among the best approaches for exploiting the spatial diversity and are considered in different standards, e.g., [15]. Then, as shown in the following, the performance of the STC-based setups is improved substantially, if they are combined with ARQ. These are the main motivations for this correspondence, in which we analyze the performance of STC-ARQ systems using adaptive power allocation 1 .