It has been reported that sarcoplasmic reticulum (SR) Ca(2+) uptake is more rapid in rat than rabbit ventricular myocytes, but little information is available on the relative SR Ca(2+) uptake activity in others species, including humans. We induced Ca(2+) transients with a short caffeine pulse protocol (rapid solution switcher, 10 mM caffeine, 100 ms) in single ventricular myocytes voltage clamped (-80 mV) with pipettes containing 100 microM fluo-3 and nominal 0 Ca(2+), in 0 Na(+)(o)/0 Ca(2+)(o) solution to inhibit Na/Ca exchange. SR in non-paced human, dog, rabbit, and mouse ventricular myocytes could be readily loaded with Ca(2+) under our experimental conditions with a pipette [Ca(2+)] = 100 nM. Resting [Ca(2+)](i) was similar in four types of ventricular myocytes. Activation of the Ca(2+)-release channel with a 100-ms caffeine pulse produced a rise in [caffeine](i) to slightly above 2 mM, the threshold for caffeine activation of Ca(2+) release. This caused a similar initial rate of rise and peak [Ca(2+)](i) in the four types of ventricular myocytes. However, there were significant differences in the duration of the plateau (top 10%) [Ca(2+)](i) transients and the time constant of the [Ca(2+)](i) decline (reflecting activity of the SR Ca(2+)-ATPase), with values for human > dog > rabbit > mouse. In paced myocytes under physiologic conditions, SR Ca(2+) content was greater in mouse than in rabbit myocytes, while peak I(Ca,L) was smaller in mouse. These findings confirm substantial species difference in SR Ca(2+)-ATPase activity, and suggest that the smaller the animal and the more rapid the heart rate, greater the activity of the SR Ca(2+)-ATPase. In addition, it appears that substantial species differences exist in the degree of SR Ca(2+) loading and I(Ca,L) under physiologic conditions.