Objective: The two-step approach recommended by the National Diabetes Data Group (NDDG), Carpenter and Coustan (C&C), and O’Sullivan, and the single-step approach recommended by the International Association of Diabetes and Pregnancy Study Group (IADPSG) are used to diagnose gestational diabetes mellitus (GDM). We aimed to determine GDM prevalence and to compare the two-step and single-step approaches used in the southeastern region of Turkey. Materials and Methods: In total, 3048 records of pregnant women screened for GDM between 2008 and 2014 were retrospectively extracted from our laboratory information system. GDM was defined according to the criteria of NDDG, C&C, and O’Sullivan between in 2008 and 2011, and according to those of the IADPSG between 2012 and 2014. Demographic variables were compared using student’s t-test. The linear trends in GDM prevalence with age were calculated using logistic regression. Results: GDM prevalence was found as 4.8%, 8%, and 13.4% using the NDDG, C&C, and O’Sullivan two-step approach, respectively, and 22.3% with the IADPSG single-step approach. GDM prevalence increased with increasing age in both approaches. Conclusion: GDM prevalence was higher using the single-step approach than with the two-step approach. There was a significant increase in GDM prevalence using the IADPSG criteria.