ALOPEX is a general optimization process incorporating a cost function containing a large number of parameters which may be simultaneously adjusted until the cost function reaches an optimum (maximum or minimum); local extremes are avoided by introducing random noise into the procedure. In this paper, ALOPEX is incorporated into a simple haemodynamic study in which an electric analogue model of the left ventricle is used to develop equations of myocardial stroke work. Pilot experiments were undertaken in rabbits (n = 5) to gauge the effectiveness of this optimizing technique. In the control state, calculated stroke work for the rabbit was determined to be 50 +/- 7 mmHg ml, while ALOPEX predicted a stroke work of 51 +/- 7 mmHg ml. ALOPEX is capable of following changing cardiovascular states when pharmacological agents are introduced. For example, after nitroprusside treatment, stroke work was reduced by 38 +/- 6% (P < 0.05) while ALOPEX predicted a 42 +/- 4% reduction from baseline (P < 0.05). Methoxamine treatment increased stroke work by 74 +/- 34%, while ALOPEX predicted a 73 +/- 43% increase above control values. There were no statistical differences between calculated and ALOPEX predicted values. Individual model parameters such as maximum left ventricular elastance (Emax) and left ventricular end diastolic volume (EDV) were also predicted correctly by ALOPEX. We have found that the ALOPEX optimization technique is useful in predicting components of multi-parametric functions. In particular, we have shown it to be adaptable to a simple haemodynamic model.