A fundamental physiological idea is that the current configuration, sets or reflects parameters within the neurophysiological feedback loops that translate task intentions into muscular activity (3, 4). Human standing requires maintenance of unstable equilibrium with minimal muscular cost while controlling configuration. Whether the postural system prioritises maintenance of joint angles at the expense of muscular costs or prioritises minimisation of joint moments while allowing joint angles to change is uncertain (1, 2). Using gentle perturbations at the knee and observation of the transition into standing from walking, recent evidence (1) shows a common tendency to stiffen the leg joints in a generalised and prevailing manner. This stiffening may result from prior settings within the postural system. This hypothesis is tested by applying an internal perturbation which requires changing joint angles to minimise active joint moments. Hypothesis: perturbation related change in active joint moments can be predicted from the preceding posture. Twelve participants stood on a force-plate holding a 5 kg bar in both hands. The bar was slowly raised to shoulder level, then slowly moved horizontally forwards to the maximum extent, then held in the forward position. Acceleration was minimal. If joint resistance were low, then during forward movement of the bar, the body should move backwards leaving the combined center of mass of body and bar in the initial position. If leg and trunk position remain constant the combined center of mass would move forwards. Summarising participants over three repetitions, in response to moving the bar anteriorly, the center of gravity moved anteriorly 26±17 mm and the absolute joint moment summed over the ankle, knee and hip increased 32±39 Nm. Both quantities showed substantial difference between participants and no change in repetition (Two-way ANOVA, α=0.05), so repetitions were averaged. Using forwards, stepwise multiple regression, the increase in summed absolute joint moment normalised by summed absolute joint rotation is predicted by the preceding knee joint extension and hip joint to gravitation vector separation (F(4,7)=14, adjusted R2=0.71, p=0.0016). The fullest model prediction (F(4,7)=40, adjusted R2=0.93, p=0.00007), confirmed trunk angle as most important. The same process applied to normalised EMG activity summed over fourteen leg muscles, also confirmed trunk angle as most important. In this task preceding posture predicts the prioritisation of force minimisation over position control. Control of trunk alignment may be hierarchically and temporally prior to the control of generalised leg stiffness.