The high-quality interpretation of seismic refraction data depends on the accurate and reliable identification of the first arrival times. First arrivals can be identified on a graphic or image by conventional picking, but this process depends on external factors, such as the scale and quality of the imaging data, amplitude ratio, sensitivity of the picking cursor and user experience. Under these considerations, identifying first arrivals in noisy data becomes more complex and unstable. In this study, the Cross-Correlation Technique (CCT), which is widely used in the process of analyzing reflection data, has been used to pick the first arrival times in noisy or noiseless seismic refraction data by a semi-automatic process. The CCT has reduced the dependence on user and decreased incorrect picking caused by environmental noise, displaying characteristics and scaling factors. The CCT has been tested with synthetic models with different noise contents and various field data. The Chi-square error criterion was used to assess the performance of the pickings. In addition, effects of small-time differences between the conventional picking process and the CCT have been demonstrated on a refraction tomography velocity section. Therefore, we believe that our proposed method is a useful contribution to the existing methods of first arrival picking.