Previous CDMA/ALOHA network performance analysis papers do not consider memory impulse channels. The Markov-Gaussian model can characterize the bursty nature of the memory impulsive noise in some wireless channels and power line communication (PLC) channels. In this paper, we propose the use of the Markov-Gaussian model in the throughput analysis of CDMA/ALOHA networks. The state transition diagram for the throughput analysis has one extra dimension to represent if the impulse noise is present or not. We derive the exact throughput analysis in memory impulse channels with and without channel load sensing protocol (CLSP). The analytic results demonstrate that if the level of the impulsive noise is higher, performance degradation is more and vice versa. Based on the analytic results, we also find the optimal CLSP threshold alpha for various impulse noise conditions. The analytic results of the throughput performance are all matched by the simulation results.