Objective: To investigate the personal features associated with dropout from regular outpatient care among persons with type 2 diabetes mellitus (DM).Methods: A total of 160 DM patients were enrolled in the study. As a retrospective analysis, outpatient's clinical characteristics, lifestyle, or social features were gathered from their medical records or interview sheets. All the subjects were divided into two groups by adherence to diabetic care, namely, 'dropout case' (DC) or 'ongoing case' (OC), and were subjected to comparative analysis. We called the patients who did not receive outpatient treatment from the clinic on a regular basis, including treatment from other clinics or dropout of diabetic care, as DC. In contrast, patients who regularly visited the clinic were defined as OC. An unconditional multiple logistic regression analysis was performed to analyze the association of adherence to diabetic care with several personal features.Results: Sixty-eight of 160 subjects (42.5%) were recognized as DC. The remaining 92 subjects (57.5%) were considered as OC. Young age (p=0.045), low plasma glucose (p=0.005) and hemoglobin A1c (HbA1c) levels (p=0.005), nonmedication (p<0.001) and no past history of DM (p=0.007) at the initial visit were the features related to dropout by crude analysis. Even after adjustment for age and gender by multivariate analysis, there remained significant inverse associations of dropout with HbA1c level, medical treatment (oral agents or insulin) and previous DM history. Neither occupation, distance from residence to clinic, smoking habit nor drinking habit was associated with dropout. Dropout mostly occurred after the initial or second visit.Conclusions: A mild condition of DM may be related to dropout from regular outpatient care. It may be necessary to clearly show the objectives and importance of regular visit to an outpatient clinic for diabetic care, particularly for screened mild DM cases in public health activities.