Detecting SGA (small for gestational age) during pregnancy improves the fetal and neonatal prognosis. To date, there is no valid antenatal biomarker of SGA used in clinical practice. Maternal circulating DLK1 (delta-like non-canonical notch ligand 1) levels have been shown to be significantly lower in pregnant women at 36 weeks of gestation (WG) who delivered a SGA newborn than in controls. Data in the literature are contradictory on the association between maternal circulating DLK1 levels and placental vascular dysfunction. The objective was to determine if maternal DLK1 levels in the second trimester of pregnancy are predictive of SGA, and to assess whether the measurement of DLK1 levels in maternal blood could be a means to distinguish SGA with placental vascular dysfunction from that due to other causes. We conducted a nested cased-control study within the EDEN mother-child cohort. 193 SGA (birth weight < 10th percentile) and 370 mother-child control (birth weight between the 25th and 75th percentile) matched pairs were identified in the EDEN cohort. Maternal circulating DLK1 levels at 26 WG were significantly lower for children born SGA than for controls (27.7 ± 8.7 ng/mL vs 30.4 ± 10.6 ng/mL, p = 0.001). Maternal blood DLK1 levels in the first quartile (DLK1 < 22.85 ng/mL) were associated with an odds ratio for SGA of 1.98 [1.15 - 3.37]. DLK1 was less predictive of SGA than ultrasound, with an area under the curve of 0.578. Maternal circulating DLK1 levels were not significantly different in cases of SGA with signs of placental vascular dysfunction (n = 63, 27.1 ± 9.2 ng/mL) than in those without placental dysfunction (n = 129, 28.0 ± 8.5 ng/mL, p = 0.53). The level of circulating DLK1 is reduced in the second trimester of pregnancy in cases of SGA at birth, independently of signs of placental vascular dysfunction. However, DLK1 alone cannot predict the risk of SGA.