In the USA and Japan, body mass index (BMI) has increased over the last several decades, whereas energy intake (EI) has decreased. However, self-reported EI data may show systematic errors. Using the calibration approach for attenuating the systematic error of self-reported EI, we aimed to compare trends in BMI and EI with and without calibration in adults from the USA and Japan. This cross-sectional study included 38,370 Americans evaluated in the National Health and Nutrition Examination Survey 2003–2018, and 200,629 Japanese evaluated in national nutrition surveys in Japan 1995–2019. EI was estimated using at least 1 day of 24-h diet recalls for Americans and 1 day of household-based dietary records for Japanese. The calibrated EI was calculated using a previously developed equation based on total energy expenditure (TEE) measured by doubly labelled water method. Using data from a review, uncalibrated EI was −20.2% and calibrated EI was −4.1% compared to the TEE; the calibration approach attenuated EI underestimation. In the USA, uncalibrated EI decreased (annual percentage change [APC]: −0.24%), but calibrated EI and BMI increased (calibrated EI, APC: 0.04%; BMI, APC: 0.32%). In Japan, the decrease was smaller for the calibrated EI than for the uncalibrated EI (uncalibrated EI, APC: −0.23%; calibrated EI, APC: −0.04%). Uncalibrated EI decreased and BMI increased in the USA and Japan, and calibrated EI increased in the USA and decreased slowly in Japan. Calibration may attenuate systematic bias in dietary assessments and facilitate the effective use of dietary data.