Expansive soils undergo alternate swelling and shrinkage respectively in rainy seasons when they absorb water and in summers when water evaporates from them. Hence, all types of foundations constructed in expansive soils are also subjected to alternate swelling and shrinkage in rainy and summer seasons. As a result, super-structure members are also affected, undergoing severe distress. Various tension-resistant and innovative foundation techniques such as belled piers and under-reamed piles have been devised for arresting the problems posed by expansive soils. Granular pile-anchor (GPA) is a recent innovative technique suggested for expansive clay beds. GPA is quite effective in controlling heave or swelling. Useful experimental heave data were obtained on laboratory scale and field scale GPAs. However, it is also necessary to study the behaviour of GPA-reinforced expansive clay beds subjected to swelling and shrinkage. This paper presents experimental data on shrinkage of GPA-reinforced expansive clay beds. Laboratory scale GPA-reinforced expansive clay beds were subjected to alternate swell–shrink cycles, each cycle for a duration of 300 days. Each clay bed was subjected to three swell–shrink cycles (N = 1, 2 and 3), each cycle monitoring swelling for 10 days and shrinkage for 90 days. The number of GPAs (n) reinforcing the clay beds was varied as 0, 1, 2 and 3. Shrinkage (mm) of a clay bed, recorded in a given swell–shrink cycle, decreased with increasing number of GPAs (n). Further, shrinkage of a given clay bed decreased significantly with increasing number of swell–shrink cycles (N) also. It was also found that shrinkage (mm) of a particular layer in the clay bed decreased with increasing depth (z) of the layer from the top of the clay bed.