The perception of a stimulus can strongly be modified by its spatial context. For example, when a vernier stimulus is flanked by two lines, performance deteriorates compared to a vernier without flanks. This is usually explained by local interactions, such as lateral inhibition or spatial pooling, between the vernier and the flanks. Here, we argue that local mechanisms are inadequate to explain this contextual modulation. In a range of experiments, we showed that performance depended on the global configuration of all stimulus elements. We presented observers with verniers flanked by lines of either same or different color, contrast polarity, or stereoscopic depth as the vernier. In all cases, performance is better when the flanks “ungroup” from the vernier. A similar effect was found when a vernier was flanked by two cuboids compared to two lines which are contained in the cuboids: Performance with flanking cuboids was superior. These results show that global aspects are crucial in contextual modulation. Vernier discrimination can be used as a tool to measure the strength of Gestalt laws