Regarding IEEE 802.11 wireless local area networks (WLANs), many researchers are focusing on signal-to-noise ratio (SNR)-based rate adaptation schemes, because these schemes have the advantage of accurately selecting transmission rates that suit the channel. However, even SNR-based rate adaptation schemes work poorly in a rapidly varying channel environment. If a transmitter cannot receive accurate rate information due to fast channel fading, it encounters continuous channel errors, because the cycle of rate adaptation and rate information feedback breaks. A well-designed request-to-send/clear-to-send (RTS/CTS) frame exchange policy that accurately reflects the network situation is an indispensable element for enhancing the performance of SNR-based rate adaptation schemes. In this paper, a novel rate adaptation scheme called adaptive RTS/CTS-exchange and rate prediction (ARRP) is proposed, which adapts the transmission rate efficiently for variable network situations, including rapidly varying channels. ARRP selects a transmission rate by predicting the SNR of the data frame to transmit when the channel condition becomes worse. Accordingly, ARRP prevents continuous channel errors through a pre-emptive transmission rate adjustment. Moreover, ARRP utilizes an efficient RTS/CTS frame exchange algorithm that considers the number of contending stations and the current transmission rate of data frames, which drastically reduces both frame collisions and RTS/CTS-exchange overhead simultaneously. Simulation results show that ARRP achieves better performance than other rate adaptation schemes.