The main goal of this study was to investigate the visual characteristics, recovery rate, and flexural properties of sawn boards from a fibre-managed plantation Eucalyptus globulus resource as a potential raw material for structural building applications. The impacts of the visual characteristics, strength-reducing features, and variation in basic density and moisture content on the bending modulus of elasticity (MOE) and modulus of rupture (MOR) of the boards were investigated. The reliabilities of different non-destructive methods in predicting MOE and MOR of the boards were evaluated, including log acoustic wave velocity measurement and numerical modellings. The MOE and MOR of the boards were significantly affected by the slope of grain, percentage of clear wood, and total number of knots in the loading zone of the boards. The normal variation in basic density significantly influenced the MOE of the boards while its effect on the MOR was insignificant. The numerical models developed using the artificial neural network (ANN) showed better accuracies in predicting the MOE and MOR of the boards than traditional multi-regression modelling and log acoustic wave velocity measurement. The ANN models developed in this study showed more than 78.5% and 79.9% success in predicting the adjusted MOE and MOR of the boards, respectively. © 2018, © 2018 Informa UK Limited, trading as Taylor & Francis Group.