The authors estimated the benchmark durations (BMDs) and their 95% lower confidence limit (BMDL) for the reference duration of shiftwork for weight gain. A 14-yr prospective cohort study was conducted in male workers at a Japanese steel company (n = 7254) who had received annual health check-ups between 1991 and 2005. The endpoints in the study were either a 5%, 7.5%, or 10% increase in body mass index (BMI) during the period of observation, compared to the BMI at entry. The association between the duration of shiftwork and weight gain was investigated using multivariate pooled logistic regression analyses with stepwise selection of covariates, including age, BMI measured during the study, drinking and smoking habits, and habitual exercise. The BMDL/BMD for shiftwork in subjects aged in their 40s or ≥50 yrs was estimated using benchmark responses (BMRs) of 5% or 10% and parameters for the duration of shiftwork and other covariates. For workers aged in their 40s, the BMDL/BMD for shiftwork with a BMR of 5% was 18.6/23.0 yrs (≥7.5%) and 16.9/19.4 yrs (≥10%). For workers aged ≥50 yrs, the BMDL/BMD with a BMR of 5% was 22.9/28.2 yrs (≥7.5%) and 20.6/23.6 yrs (≥10%). The reference duration of shiftwork that associated with weight gain was shown to be at least 17 yrs in middle-aged workers. Special attention should be paid to prevent weight gain at an earlier stage and not when this increase in weight has become apparent.