As far as the aerodynamic characterization of a flying vehicle is concerned, flight testing is probably the most accurate approach as it perfectly resembles the real flight environment. Flight data are obtained by tracking the vehicle via radars if modifying the vehicle design is not recommended/attainable. In the open literature, different techniques are used to analyze radar data; the key issue is the computational demands of each technique and the quality of the resulting aerodynamic characteristics. In this paper, three techniques are considered namely, Least-Square (LS), Maximum-Likelihood Estimation (MLE), and Stepwise Regression (SR), with focus on the prediction of the drag coefficient of a case study vehicle. Features for each technique are addressed based on brief previous published data. A new variant of the MLE method is proposed based on the physical segmentation of the available dataset. Predicted point-mass trajectories are compared with own comprehensive flight test to assess the techniques in concern. It is concluded that Stepwise-regression outperforms with a large dataset, while Maximum-Likelihood Estimation is more feasible considering the lack of data. The proposed variant of the MLE method yields more accurate drag prediction compared to the basic one.