Vitamin A deficiency (VAD) is a significant public health problem in many countries. While cost-effective interventions are available to control VAD, reliable information is needed to the track progress of control programmes. However, assessment of VAD is uncommon because current approaches are expensive and not feasible in low-resource settings. The present study explores the utility of retinol-binding protein (RBP), analysed by enzyme-linked immunosorbent assay from capillary blood, as an alternative measure of serum retinol concentrations in populations. The study collected matched panels of venous and capillary blood from pre-school children in Chiang Mai, Thailand. Of a total sample of 195 children, there were no differences between RBP from venous blood, RBP from capillary blood or retinol from capillary blood relative to retinol from venous blood. Receiver-operating characteristic curve analysis suggested a cut-off of RBP < 0.825 μmol l−1 had optimal screening proficiency relative to retinol <0.70 μmol l−1. For the purpose of population assessment, all three parameters performed well in screening for VAD relative to retinol from venous blood. There were no differences in the estimates of VAD between children stratified by inflammation status. Lower RBP concentrations were found in children in the early convalescent stage of infection than in children with no infection or in the late convalescent stage. This study provided evidence of the biological comparability between retinol and RBP estimated from venous blood and capillary blood. This is a critical observation as it provides empirical evidence that RBP from capillary blood is a surrogate measure of serum retinol concentrations.