Although accurate assessment of energy intake (EI) is critical in diabetes care, underreporting of EI on dietary records (DR) is often an issue. However, few studies have examined EI with doubly labeled water (DLW) in patients with diabetes mellitus. We aimed to investigate the impact of sex and obesity on the dissociation of DR from total energy expenditure (TEE) evaluated with DLW in patients with type 2 diabetes.Fifty-two patients with type 2 diabetes aged 60-79 years were enrolled for the Clinical Evaluation of Energy Requirements in Patients with Diabetes Mellitus (CLEVER-DM) study at a single university hospital. TEE was measured over 14 days by the DLW method as standard. EI was calculated by assessment of 3-day DR by registered dietitians.The mean difference between EI and TEE was 238 ± 412 kcal/day (~10% of TEE). Neither EI nor TEE was significantly different between obese (body mass index (BMI) ≥25 kg/m2) and non-obese (BMI <25 kg/m2) patients. There was a negative correlation between EI/TEE ratio and BMI in women (R = -0.437, P = 0.033) but not in men (R = -0.174, P = 0.377). There was a significant difference in EI/TEE ratio between obese and non-obese patients among women (0.85 ± 0.15 vs. 1.01 ± 0.21, P = 0.045) but not men (0.85 ± 0.20 vs. 0.87 ± 0.17, P = 0.79).EI calculated by 3-day DR may underestimate habitual intake, which is assumed to be equal to TEE measured by the DLW method except in non-obese women with diabetes.UMIN000023051.