Background There are limited data on the role of body image in patients with type 2 diabetes. The purpose of this study was to compare body self-esteem in this group with norms for the general Polish population and to investigate the relationship between body self-esteem and the psychological and clinical characteristics of the course of diabetes. Methods A group of 100 consecutive adult patients with type 2 diabetes (49 women and 51 men) aged 35 to 66 years were assessed using the Body Esteem Scale (BES), World Health Organization-Five Well-Being Index (WHO-5), Problem Areas in Diabetes Scale (PAID), and Hamilton Rating Scale for Depression (HAM-D). Results In comparison to norms for the general population, women with type 2 diabetes had lower body self-esteem only in the dimension of Physical Condition (M = 30.71; SD = 7.11 versus M = 32.96; SD = 5.69; P = 0.003), whereas men in the dimensions of Physical Condition (M = 42.43; SD = 9.43 versus M = 48.30; SD = 8.42; P <0.001) and Upper Body Strength (M = 32.16; SD = 6.60 versus M = 33.97; SD = 5.86; P = 0.015). There were moderate or weak positive correlations between the overall BES score and/or its dimensions and subjective well-being, and negative correlations between the overall BES score and/or its dimension and the severity of depression symptoms, level of glycated hemoglobin (HbA1c), body mass index (BMI), and diabetes-related distress among women. Among men, BES scores were positively correlated with well-being, and negatively, with BMI and diabetes-related distress. A correlation of r = 0.39 between BES scores and HbA1c levels was relatively high compared with values for other psychosocial factors. Both in women and men, a high Physical Condition score was a significant predictor of better well-being, less severe depression, and milder diabetes-related distress. Among men, it was also a significant predictor of lower BMI, whereas among women, BMI was predicted by Weight Concern. Conclusions Persons with diabetes seem to have lower body self-esteem than the general population, which is significantly associated with clinical and psychological characteristics of the diabetes course. The observed differences and relationships are gender-specific.