Centrifuge modelling is an advantageous technique to simulate the behaviour of helical anchors in sand. The uplift capacity of a single-helix anchor is the sum of the helix bearing resistance plus the shaft resistance; therefore, recommendations to avoid the grain-size effect of previous model tests on plate anchors and on regular piles appear to be reasonable for helical anchor model tests. However, there are two important differences between centrifuge models of plate anchors and helical anchors: the ratio of the helical plate to the shaft diameter and the plate installation procedure. Consequently, the shaft friction and the plate bearing resistance of helical anchors can be affected by the scaling effect in a different manner. To clarify this uncertainty, this study was proposed to investigate the grain-size effect on the results of uplift capacity of helical anchor models in sand. Anchor models with different scales were tested to simulate the same prototype in very dense dry sand in a centrifuge. These experiments have shown that no scale effect was found for models with the ratio of the effective helical radius to the average grain size >58. The findings obtained here could guide future centrifuge model studies on helical anchors.