We reviewed the current literature on perioperative 'optimization' of the circulation and the circulatory resuscitation goals in critically ill patients after (non-cardiac) surgery, sepsis or trauma, with the help of variables obtained by a pulmonary artery catheter. Optimization goals that include a maximal stroke work to filling pressure relationship, an elevated (supranormal) cardiac output (> 4.5 l/min/m2), an elevated O2 delivery (> 600 ml/min/m2) and uptake (> 150-170 ml/min/m2), may ameliorate or prevent a tissue O2 deficit and may be associated with less organ failures and improved survival in patients with or at risk for shock. However, this policy may not benefit patients with less severe circulatory deterioration or preterminal illness, where an increase in O2 delivery may only load the heart and may not increase systemic O2 uptake. In fact, in the course of trauma or sepsis, boosting O2 delivery may not increase O2 uptake and thereby decrease mortality, if prevented by a severe O2 extraction deficit (high mixed venous O2 saturation) because of maldistribution of blood flow associated with vasodilatation. The latter may prevent, in turn, an improvement in (regional) markers of perfusion inadequacy, including oliguria, a raised gastric to blood PCO2 gradient (tonometry) and lactic acidaemia. In contrast, blocking β-receptors prior to surgery in patients at high risk for cardiac events, rather than for multiple organ failure, may decrease morbidity and mortality after major surgery. The doubt cast on the contribution of the pulmonary artery catheter to the management of critically ill patients has been fed, among others, by studies disproving the survival benefit of the above optimization goals. It is likely that future studies on circulatory optimization goals will include both global and regional tissue perfusion/oxygenation variables, mean arterial blood pressure and heart and renal function. A study on a set of variables to guide treatment is probably hard to perform, but more likely to help defining the usefulness of circulatory optimization, than a study on the value of manipulating only one variable at a time.