ABSTRACT Objective Detection of fetal growth restriction (FGR) remains poor and most screening strategies rely on cross‐sectional evaluation of fetal size during the third trimester. A longitudinal and individualized approach has been proposed as an alternative method of evaluation. The aim of this study was to compare second‐ to third‐trimester longitudinal growth assessment to cross‐sectional evaluation in the third trimester for the prediction of small‐for‐gestational age (SGA) and late FGR in low‐risk singleton pregnancy. Methods This was a prospective cohort study of 2696 unselected consecutive low‐risk singleton pregnancies scanned at 21 ± 2 and 32 ± 2 weeks. For cross‐sectional growth assessment, abdominal circumference (AC) measurements were transformed to z‐values according the 21st‐INTERGROWTH standards. Longitudinal growth assessment was performed by calculating the AC z‐velocity and the second‐ to third‐trimester AC conditional growth centile. Longitudinal assessment was compared with cross‐sectional assessment at 32 weeks. Association of cross‐sectional and longitudinal evaluations with SGA and late FGR was assessed by logistic regression analysis. Predictive performance was determined by receiver–operating characteristics curve analysis. Result In total, 210 (7.8%) newborns were classified as SGA and 103 (3.8%) as late FGR. Neither longitudinal measurement improved the association with SGA or late FGR provided by cross‐sectional evaluation of AC z‐score at 32 weeks. Areas under the curves of AC z‐velocity and conditional AC growth were significantly smaller than those of cross‐sectional AC z‐scores (P < 0.001), although AC z‐velocity performed significantly better than did conditional AC growth (P < 0.001). Conclusion Longitudinal assessment of fetal growth from the second to third trimester has a low predictive capacity for SGA and late FGR in low‐risk singleton pregnancy compared with cross‐sectional growth evaluation. Copyright © 2017 ISUOG. Published by John Wiley & Sons Ltd.