We examined the reproducibility of dual-energy x-ray absorptiometry (DXA) visceral adipose tissue (VAT) estimates as well as the impact of pre-assessment diet. In a counterbalanced design, 41 adults received 6 DXA scans, while consuming standardized high- and low-carbohydrate (LC) diets. The impact of pre-assessment diet was examined via repeated-measures analysis of variance and reproducibility of VAT estimates was evaluated by technical error of measurement (TEM) and s.e. of a single determination (SESD). VAT area, mass and volume were acutely decreased by ~ 6.5% after the LC diet (P ≤ 0.001), but not the high-carbohydrate diet (P > 0.3). Differences persisted after an overnight fast. TEMs for VAT area, mass and volume were < 5 cm2, ≤ 0.2 kg and < 25 cm3. SESDs for all VAT variables were ≤ 5.5%. In conclusion, DXA VAT estimates exhibit good reproducibility in controlled conditions, but may be impacted by pre-assessment diet.