We investigated left ventricular (LV) early diastolic filling in 10 normal conscious dogs that had been previously instrumented to measure LV and left atrial (LA) pressures and three orthogonal LV internal dimensions. LV volume was calculated as a general ellipsoid. The pressure within a passive structure increases as it is filled. If myocardial relaxation is rapid enough to substantially aid LV diastolic filling, it may overcome this effect and cause LV pressure to fall despite an increase in volume. Thus, we defined the amount of LV filling that occurred while LV pressure was falling as relaxation filling, which is a measure of the importance of LV relaxation during early diastolic filling. The time constant of relaxation (T) was derived from the exponential fall of LV pressure during isovolumic relaxation. While LV pressure was falling early in diastole (the relaxation filling period), all three LV diameters increased. Autonomic blockade with hexamethonium (5 mg/kg) and atropine (0.1 mg/kg) reduced relaxation filling from 21 +/- 6% (mean +/- SD) to 12 +/- 3% of the stroke volume (p less than 0.01). The mean LA pressure also was significantly decreased (from 12 +/- 2 to 10 +/- 5 mm Hg, p less than 0.05), while the duration of the relaxation filling period and T were unchanged. Positive inotropic stimulation with dobutamine (10 micrograms/kg/min) shortened T without changing LA pressure. The maximum LA-LV pressure gradient, dV/dtmax, and relaxation filling all increased. Augmented preload produced by dextran infusion (500 ml/10 min) caused an increase in LA pressure (from 11 +/- 3 to 21 +/- 8 mm Hg, p less than 0.05) without altering T. This also increased the maximum LA-LV pressure gradient, dV/dtmax, and relaxation filling. Augmented afterload produced by methoxamine (10 mg/3 min i.v.) significantly increased LA pressure (from 9 +/- 4 to 15 +/- 10 mm Hg, p less than 0.05) and lengthened T (from 35 +/- 4 to 50 +/- 7 msec, p less than 0.05) and the duration of relaxation filling (from 36 +/- 5 to 44 +/- 9 msec, p less than 0.01) without altering the maximum LA-LV pressure gradient, dV/dtmax, or LV relaxation filling. Incremental changes in heart rate induced by atrial pacing (from 100-180 beats/min) resulted in progressive decreases in the time constant of LV relaxation and the duration of relaxation filling. The LA pressure was also decreased. There was no corresponding increase in the amount of active LV filling until the heart rate reached 180 beats/min. During all these interventions, T correlated with the duration of LV relaxation filling (r = 0.99. p less than 0.05). The amount of relaxation filling and dV/dtmax both correlated with the maximum LA-LV pressure gradient.(ABSTRACT TRUNCATED AT 400 WORDS)