The effects of left ventricular diastolic pressure changes on the values of the time constant (T) of isovolumic pressure fall are controversial. Normally, T is calculated either by linear regression of 1n left ventricular pressure (LVP) vs. time (TL) or by using an exponential model with asymptote (PB, extrapolated base-line pressure to which LVP would fall if decay continued indefinitely). This study, in intact dogs, has been designed to revise the effects that drugs that alter load (angiotensin and nitroprusside, without and with autonomic blockade) and inotropism (isoproterenol and propranolol) might have on the relaxation rate using different numerical methods. We found that 1) the upward or downward translation of the LVP curve had an important effect on the value of TL calculated by the semilogarithmic method; 2) when a model with asymptote was used, the simultaneous changes of T and PB, were not related to the dose of the drug; 3) the values of TL and the -dP/dt values extrapolated to 15 mmHg from a model with PB, were much more sensitive to beta-agonist or antagonist drugs than to others whose action was through load alteration. We feel that some of the earlier studies on relaxation rate determinants should be revised as possibly some of the conclusions have been based on artifacts introduced by the method chosen for the computation of the relaxation parameters.