Scaled sandbox experiments with high basal friction, simulating the growth of accretionary wedges, display cycles alternating between frontal imbricate thrusting and underthrusting of long, undeformed sheets. By contrast, low basal friction experiments with otherwise similar and constant, initial conditions produce a classic frontal imbricate fan through repeated failure along frontal thrusts. The cyclical behavior observed in high basal friction experiments is expressed by three quantities: (1) the average spacing between frontal thrusts, (2) the advance and retreat of the deformation front, and (3) the frontal slope (Alpha) of the actively deforming wedge. As a long sheet is underthrust, the front is steepened through slumping until the maximum critical angle is reached. Then frontal thrusting resumes and the accretion of imbricate slices builds the wedge forward, thereby lowering the taper to the minimum critical angle. At shallow tapers, a long unit is underthrust and subsequently uplifts, shortens, and steepens the overlying wedge through backthrust deformation, thus completing the cycle. Underthrusting of long units offers a simple mechanism for underplating overlying units. It also provides a possible explanation for temporally and spatially varying wedge geometries in nature, when basal frictions attain 80%–90% of the internal friction.