The effect of buoyant forces on the motion of a large object immersed in a bubbling fluidized bed (BFB) was experimentally studied using digital image analysis. The experiments were performed in a 2-D bubbling fluidized bed with glass spheres as bed material and cylindrical objects with different densities and sizes. The object motion was measured using non-intrusive tracking techniques. The effect of gas velocity was also analyzed. The circulation of an object in a BFB is defined by several parameters. The object might be able to circulate homogeneously throughout the bed or stay in preferred regions, such as the splash zone or the bottom zone. While circulating, the object moves back and forth between the surface of the bed and the inner regions, performing a series of cycles. Each cycle is composed by sinking and rising paths, which can be one or several, depending on whether a passing bubble is able to lift the object to the surface or the object is detached from it or its drift at an intermediate depth. Therefore, the number of rising paths or number of jumps that the object undergo in a cycle, interleaved with sinking paths, and the maximum attained depth characterize each cycle, together with the mean sinking and rising velocities of the object. In this work, experimental measurements of the probability distributions of the number of jumps and the maximum attained depth, the axial homogeneity of object motion and rising and sinking object velocities are presented for objects with different sizes and densities. The results show a coherent behavior, independent of density and size, for the probability distributions of the number of jumps. This is also true for the maximum attained depth, but only when a proper circulation throughout the bed is ensured. Such a proper circulation and axial homogeneity is, on the other hand, much affected by object density, size and gas velocity. Rising and sinking velocities are highly dependent on gas velocity, as established in well-known models of bubble and dense phase velocities. Nevertheless, rising velocities are practically unaffected by object density or size, while sinking velocities show a low dependence on density and a steeper one on size. These results suggest that buoyant forces are relevant during the sinking process, and almost neutral during the rising path. (C) 2011 Elsevier Ltd. All rights reserved.