Small for gestational age (SGA) is defined by weight (in utero estimated fetal weight or birth weight) below the 10th percentile (professional consensus). Severe SGA is SGA below the third percentile (professional consensus). Fetal growth restriction (FGR) or intra-uterine growth restriction (IUGR) usually correspond with SGA associated with evidence indicating abnormal growth (with or without abnormal uterine and/or umbilical Doppler): arrest of growth or a shift in its rate measured longitudinally (at least two measurements, 3 weeks apart) (professional consensus). More rarely, they may correspond with inadequate growth, with weight near the 10th percentile without being SGA (LE2). Birthweight curves are not appropriate for the identification of SGA at early gestational ages because of the disorders associated with preterm delivery. In utero curves represent physiological growth more reliably (LE2). In diagnostic (or reference) ultrasound, the use of growth curves adjusted for maternal height and weight, parity and fetal sex is recommended (professional consensus). In screening, the use of adjusted curves must be assessed in pilot regions to determine the schedule for their subsequent introduction at national level. This choice is based on evidence of feasibility and the absence of any proven benefits for individualized curves for perinatal health in the general population (professional consensus). Children born with FGR or SGA have a higher risk of minor cognitive deficits, school problems and metabolic syndrome in adulthood. The role of preterm delivery in these complications is linked. The measurement of fundal height remains relevant to screening after 22 weeks of gestation (Grade C). The biometric ultrasound indicators recommended are: head circumference (HC), abdominal circumference (AC) and femur length (FL) (professional consensus). They allow calculation of estimated fetal weight (EFW), which, with AC, is the most relevant indicator for screening. Hadlock's EFW formula with three indicators (HC, AC and FL) should ideally be used (Grade B). The ultrasound report must specify the percentile of the EFW (Grade C). Verification of the date of conception is essential. It is based on the crown-rump length between 11 and 14 weeks of gestation (Grade A). The HC, AC and FL measurements must be related to the appropriate reference curves (professional consensus); those modelled from College Francais d'Echographie Fetale data are recommended because they are multicentere French curves (professional consensus). Whether or not a work-up should be performed and its content depend on the context (gestational age, severity of biometric abnormalities, other ultrasound data, parents' wishes, etc.) (professional consensus). Such a work-up only makes sense if it might modify pregnancy management and, in particular, if it has the potential to reduce perinatal and long-term morbidity and mortality (professional consensus). The use of umbilical artery Doppler velocimetry is associated with better newborn health status in populations at risk, especially in those with FGR (Grade A). This Doppler examination must be the first-line tool for surveillance of fetuses with SGA and FGR (professional consensus). A course of corticosteroids is recommended for women with an FGR fetus, and for whom delivery before 34 weeks of gestation is envisaged (Grade C). Magnesium sulphate should be prescribed for preterm deliveries before 32-33 weeks of gestation (Grade A). The same management should apply for preterm FGR deliveries (Grade C). In cases of FGR, fetal growth must be monitored at intervals of no less than 2 weeks, and ideally 3 weeks (professional consensus). Referral to a Level IIb or III maternity ward must be proposed in cases of EFW <1500g, potential birth before 32-34 weeks of gestation (absent or reversed umbilical end-diastolic flow, abnormal venous Doppler) or a fetal disease associated with any of these (professional consensus). Systematic caesarean deliveries for FGR are not recommended (Grade C). In cases of vaginal delivery, fetal heart rate must be monitored continuously during labour, and any delay before intervention must be faster than in low-risk situations (professional consensus). Regional anaesthesia is preferred in trials of vaginal delivery, as in planned caesareans. Morbidity and mortality are higher in SGA newborns than in normal-weight newborns of the same gestational age (LE3). The risk of neonatal mortality is two to four times higher in SGA newborns than in non-SGA preterm and full-term infants (LE2). Initial management of an SGA newborn includes combatting hypothermia by maintaining the heat chain (survival blanket), ventilation with a pressure-controlled insufflator, if necessary, and close monitoring of capillary blood glucose (professional consensus). Testing for antiphospholipids (anticardiolipin, circulating anticoagulant, anti-beta2-GP1) is recommended in women with previous severe FGR (below third percentile) that led to birth before 34 weeks of gestation (professional consensus). It is recommended that aspirin should be prescribed to women with a history of pre-eclampsia before 34 weeks of gestation, and/or FGR below the fifth percentile with a probable vascular origin (professional consensus). Aspirin must be taken in the evening or at least 8h after awakening (Grade B), before 16 weeks of gestation, at a dose of 100-160mg/day (Grade A).