Severely obese patients undergoing bariatric surgery (BS) are at increased risk for venous thromboembolism (VTE). How standard low molecular weight heparin (LMWH) regimen should be adapted to provide both sufficient efficacy and safety in this setting is unclear. We aimed to compare the influence of four body size descriptors (BSD) on peak anti-Xa levels in BS obese patients receiving LMWH fixed doses to identify which one had the greatest impact. One hundred and thirteen BS obese patients [median body mass index (BMI), 43.3 kg/m2 (IQR, 40.6–48.7 kg/m2)] receiving subcutaneous dalteparin 5000 IU twice daily were included in this prospective monocenter study. Peak steady-state anti-Xa levels were measured peri-operatively following thromboprophylaxis initiation. Only 48% of patients achieved target anti-Xa levels (0.2–0.5 IU/ml). In univariate analysis, age, gender, total body-weight (TBW), lean body-weight (LBW), ideal body-weight (IBW), BMI and estimated glomerural filtration rate (eGFR) were associated with anti-Xa levels. The strongest negative association was observed with LBW (r = −0.56, p < .0001). Receiver operating characteristic curves indicated that among BSD, LBW (cut-off >55.8 kg) had the highest sensitivity (73%) and specificity (69%) to predict sub-prophylactic anti-Xa levels. In multivariate analysis, LBW and eGFR remained associated with anti-Xa levels (β = −0.47 ± 0.08, p < .0001 and β = −0.19 ± 0.08; p = .02, respectively). In BS morbidly obese patients receiving LMWH for thromboprophylaxis after BS, LBW and eGFR are the main determinants of anti-Xa level, and could be proposed in LMWH-based thromboprophylaxis dosing algorithms. The efficacy of a LBW-scale based dosing algorithm for optimal VTE prevention deserves further prospective randomized trials.