Screening for gestational diabetes mellitus (GDM) during pregnancy is cumbersome. Measurement of plasma fructosamine may help simplify the first step of detecting GDM. We aimed to assess the predictive value of mid-pregnancy fructosamine for GDM, and its association with postpartum glycemic indices. Among 1488 women from Project Viva (mean ± SD: 32.1 ± 5.0 years old; pre-pregnancy body mass index 24.7 ± 5.3 kg/m²), we measured second trimester fructosamine and assessed gestational glucose tolerance with a 50 g glucose challenge test (GCT) followed, if abnormal, by a 100 g oral glucose tolerance test (OGTT). Approximately 3 years postpartum (median 3.2 years; SD 0.4 years), we measured maternal glycated hemoglobin ( = 450) and estimated insulin resistance (HOMA-IR; = 132) from fasting blood samples. Higher glucose levels 1 h post 50 g GCT were associated with higher fructosamine levels (Pearson's r = 0.06; = 0.02). However, fructosamine ≥222 µmol/L (median) had a sensitivity of 54.8% and specificity of 48.6% to detect GDM (area under the receiver operating characteristic curve = 0.52); other fructosamine thresholds did not show better predictive characteristics. Fructosamine was also weakly associated with 3-year postpartum glycated hemoglobin (per 1 SD increment: adjusted β = 0.03 95% CI [0.00, 0.05] %) and HOMA-IR (per 1 SD increment: adjusted % difference 15.7, 95% CI [3.7, 29.0] %). Second trimester fructosamine is a poor predictor of gestational glucose tolerance and postpartum glycemic indices.