Pregnant women with diabetes mellitus (DM) are at high risk for hypertensive disorder of pregnancy (HDP). Women with poor control DM sometimes have heavy-for-dates infants. However, women with HDP sometimes have light-for-dates infants. We aim to clarify the relationship between glycemic control and fetal growth in women with DM and/or subsequent HDP. Of 7893 women gave singleton birth at or after 22 gestational weeks, we enrolled 154 women with type 1 DM (T1DM) or type 2 DM (T2DM) whose infants did not have fetal abnormalities. Among women with T1DM or T2DM, characteristics of the three groups (with HDP, without HDP, and with chronic hypertension [CH]) were compared. No women with T1DM had CH, but 19 (17.4%) of 109 with T2DM did. HDP incidence was similar between women with T1DM (22.2%) and T2DM without CH (16.7%). Among women with T1DM, the incidences of fetal growth restriction (FGR) with and without HDP were similar. However, among women with T2DM without CH, this incidence was significantly higher among those with HDP (33.3%) than among those without HDP (5.3%), was significantly more common with HbA1c levels at first trimester ≥ 7.2% (33.3%) than with those < 7.2% (5.6%), and significantly more numerous without pre-pregnancy therapies for DM (23.3%) than with them (3.3%). Among women with T2DM and HDP, those with FGR had smaller placenta SDs and higher insulin dosages at delivery than those without light-for-dates. In multivariate analysis, the presence of diabetic nephropathy was a predictor of T1DM and HDP (P = 0.0105), whereas HbA1c levels ≥ 7.2% before pregnancy was a predictor of T2DM and HDP (P = 0.0009). Insulin dosage ≥ 50U/day at delivery (P = 0.0297) and the presence of HDP (P = 0.0116) independently predicted T2DM, HDP, and FGR development. Insufficient pre-pregnancy treatment of DM increased the risk of HDP.