Objective: Although the use of dipeptidyl peptidase-4 (DPP-4) inhibitors has been increasing after their first approval in 2006, little is known about their prescribing pattern. Therefore, the objective of this study is to evaluate the prescribing pattern of the DPP-4 inhibitors for the treatment of type 2 diabetes mellitus (T2DM) and examine sociological factors associated with physician prescribing behavior in the U.S. outpatient setting. Methods: This cross-sectional study was conducted utilizing data from the 2006-2010 National Ambulatory Medical Care Survey (NAMCS) and employed the Eisenberg model that explains physician decision making in the context of sociologic influences. For independent variables, the following characteristics were determined based on the Eisenberg model: patient characteristics, physician characteristics, the physician-health care system interaction, and the physician-patient relationship. The dependent variable was the use of DPP-4 inhibitors. Multivariate logistic regressions were used for analyses. Results: The estimated population size was 535,158,796 patients during five years, and 3.85% of them were prescribed DPP-4 inhibitors. Among the patient characteristic-related factors, the odds of the use of DPP-4 inhibitors was 73% lower in patients with Medicaid compared to patients with private insurance (OR = 0.27; 95% CI, 0.08-0.88; p = .030). For the physician characteristic-related factor, the odds of prescribing DPP-4 inhibitors for primary care physicians are about 86% higher than the odds for non-primary care physicians (OR = 1.86; 95% CI, 1.17-2.95; p = .008). In addition, physicians in private offices were 3.01 times more likely to prescribe DPP-4 inhibitors than physicians in the health maintenance organizations (HMO) (OR = 3.01; 95% CI, 1.03-8.78; p = .043). Conclusions: Patient characteristics, physician characteristics, and the physician’s relationship with the health care system were associated with an increased use of DPP-4 inhibitors. However, the physician’s relationship with the patient was not associated with an increased use of DPP-4 inhibitors.