In this study, the performance of a secondary network in cognitive radio (CR) is studied in the context of vehicle-to-everything (V2X). The non-orthogonal multiple access (NOMA) is effectively applied in this new system model, namely CR-assisted NOMA-V2X, and it is beneficial to serve group of vehicles. In our proposed system, two schemes related to vehicle-to-vehicle (V2V) transmissions are further considered to enhance performance of the vehicle that needs higher quality of service (QoS). However, the degradation performance can be predicted by evaluating downlink under impacts from interference from the primary network, imperfect channel state information (CSI) and imperfect successive interference cancellation (SIC). The outage performance gap among two vehicles exists since different power allocation factors were assigned to them. To validate the system performance, the outage probability is first derived in exact and approximate forms and then the throughput can be further achieved. The optimal throughput can be obtained by numerical simulations. Simulation results are provided to verify the correctness of the derived expressions and it exhibits advantages of the proposed CR-assisted NOMA-V2X system in terms of outage probability and the throughput.