This article introduces a low-complexity ultrawideband quasi-deterministic ray tracing (QD-RT) method for statistical analysis of wireless channels. This model uses a statistical distribution to model the bistatic radar cross section (RCS) of irregular objects, such as cars and pedestrians, instead of a deterministic propagation model, i.e., applying the exact values of bistatic RCSs. It is shown that the quasi-deterministic propagation model benefits from a low complexity compared with a deterministic model while keeping the accuracy. The proposed QD-RT method is applied in a realistic street canyon scenario in the millimeter-wave (mmWave) frequency band, and the performance of the QD-RT method is verified by the deterministic propagation method, where the second-order statistics, including root-mean-square (rms) delay spread and angular spread, and the first-order statistic transfer function yield good agreements. Finally, the application of the QD-RT in stochastic channel modeling is demonstrated by developing a 3rd generation partnership project (3GPP)-like statistical channel model for street canyon scenarios.