The granular pile-anchor (GPA) foundation system, a more recent technique, has also been found to be quite successful in controlling heave of expansive clay beds. This paper presents experimental data obtained from laboratory-scale heave tests conducted on an unreinforced expansive clay bed (n = 0) and expansive clay beds reinforced by a single GPA (n = 1), twin GPAs (n = 2) and a group of GPAs laid in equilateral triangular pattern (n = 3). The thickness of all the test clay beds was 200 mm. Heave was monitored at different depths of clay beds as they were inundated. Ground heave (or surface heave) was also monitored at different radial distances (r) from the centre of the test tank. The unreinforced clay bed and the GPA-reinforced clay beds were identical with regard to thickness (H), placement water content (w%) and dry density (γd). Heave decreased at all depths of the clay beds as the number of the GPAs (n) increased. Furthermore, heave decreased with increasing depth (z) from the top of the clay beds. Heave was measured to be 56·5, 48, 28·5 and 24 mm at the top of the clay bed, respectively, for n = 0 (unreinforced expansive clay bed), n = 1, 2 and 3, indicating increased resistance to heave with increasing number of GPAs. The factor of safety of the GPA foundation system against uplift also increased with increasing number of GPAs.