The main purpose of the present study has been to find an answer to the question: Can the subject generate anticipatory postural adjustments (APAs) when a predictable postural perturbation occurs in the absence of a voluntary action? Answering this question would allow us to distinguish between two competing hypotheses on the relation between APAs and voluntary movements. One hypothesis considers both APAsigma and voluntary "focal" movements different peripheral patterns associated with a single control process, while the alternative hypothesis considers them outcomes of two parallel control processes. Healthy subjects performed series of loading and unloading trials that included: (1) catching a falling load onto another load held in extended hands; (2) catching a falling load onto a tray attached to the trunk; (3) allowing a falling load to hit another load out of the extended hands, causing an unloading; and (4) releasing a load held in extended hands by a voluntary shoulder movement. In series 1, precautions were taken to avoid possible small hand movements prior to the impact of the falling load. Available visual information on the trajectory of the falling load was manipulated. In all conditions, except when the subject's eyes were closed, APAs were seen with patterns that were adequate for counteracting expected perturbations. Quantitative electromyographic indices of APAs depended on the availability of visual information and particular methods of introducing postural perturbations despite the fact that the magnitude of the perturbation was always the same. Our findings support a hypothesis that control processes resulting in APAs can be different from control processes associated with focal voluntary movements.