This project aimed to analyze the impact of disease acceptance and selected demographic and clinical factors on the adherence to treatment recommendations in elderly type 2 diabetes mellitus patients. The observational study was performed using standardized research questionnaires: the Acceptance of Illness Scale (AIS), the Self-Care of Diabetes Inventory (SCODI), and the Adherence in Chronic Diseases Scale (ACDS). Two hundred patients with T2DM were studied (age M = 70.21 years, SD = 6.63 years). The median degree of disease acceptance was 29 (min–max = 8–40) and the median level of adherence was 24 (min–max = 13–28). Disease acceptance was a significant (p = 0.002) independent predictor of the odds of qualifying for non-adherence OR = 0.903, 95% CI = 0.846–0.963. The respondents gave the lowest scores for glycemic control (Mdn = 38.99, min–max = 8.33–150), and health control (Mdn = 55.88, min–max = 11.76–100). A one-way ANOVA showed that the non-adhering patients were significantly older compared to the adherence group and were taking significantly more diabetes pills per day. The level of disease acceptance was average, but it turned out to be an independent predictor of adherence. Therefore, it is justified to use psychological and behavioral interventions that are aimed at increasing the level of diabetes acceptance in elderly people with T2DM. It is important to have a holistic approach to the patient and to take actions that consider the patient’s deficits in the entire biopsychosocial sphere. The obtained result confirmed the legitimacy of interventions aimed at increasing the level of disease acceptance in this group of patients.