Objective The purpose of this study was to evaluate the association between starchy vegetable consumption and subgroup consumption in the first trimester and the risk of gestational diabetes mellitus (GDM). Methods A prospective study (n = 1444) was conducted in China. Dietary information was assessed by 24-hour dietary recalls for three days and then we calculated the consumption of total starchy vegetable and its subgroups, including (1) potato and (2) other starchy vegetable (pumpkin, lotus root, yam, taro, water chestnut, pea, and cowpea). GDM was diagnosed according to the results of 75-g two-hour oral glucose tolerance test (OGTT) at 24–28 weeks of gestation. A modified log-binomial regression was used to estimate RRs and 95% CIs of GDM risk. Results Among the 1444 participants in our study, 520 were diagnosed with GDM. The adjusted RRs (95% CIs) for GDM from the lowest to the highest quartiles of total starchy vegetable consumption were 1.00 (reference), 1.29 (1.06, 1.57), 1.13 (0.93, 1.40), and 1.26 (1.02, 1.56), respectively; p for trend = .032. For potato, the RR of GDM risk was 1.32 for the highest potato intake quartile compared with the lowest quartile (95% CI 1.07–1.64, p for trend = .003). In addition, we did not observe an association between other starchy vegetable intakes and GDM risk. Conclusions A higher consumption of total starchy vegetables and potatoes in the first trimester is associated with a greater risk of GDM.