We describe the feasibility of chronic measurement of cardiac output (CO) in conscious mice. With the use of gas anesthesia, mice >30 g body wt were instrumented either with transit-time flow probes or electromagnetic probes placed on the ascending aorta. Ascending aortic flow values were recorded 6-16 days after surgery when probes had fully grown in. In the first set of experiments, while mice were under ketamine-xylazine anesthesia, estimates of stroke volume (SV) obtained by the transit-time technique were compared with those simultaneously obtained by echocardiography. Transit-time values of SV were similar to those obtained by echocardiography. The average difference +/- SD between the methods was 2 +/- 7 microl. In the second set of studies, transit-time values of CO were compared with those obtained by the electromagnetic flow probes. In conscious resting conditions, estimates +/- SD) of cardiac index (CI) obtained by the transit-time and electromagnetic flow probes were 484 +/- 119 and 531 +/- 103 ml x min(-1) x kg body wt(-1), respectively. Transit-time flow probes were also implanted in mice with a myocardial infarction (MI) induced by ligation of a coronary artery 3 wk before probe implantation. In these MI mice (n = 7), average (+/- SD) resting and stimulated (by volume loading) values of CO were significantly lower than in noninfarcted mice (n = 15) (resting CO 16 +/- 3 vs. 20 +/- 4 ml/min; stimulated CO 20 +/- 5 vs. 26 +/- 6 ml/min). Finally, using transfer function analysis, we found that, in resting conditions for both intact and MI mice, spontaneous variations in CO (> 0.1 Hz) were mainly due to those occurring in SV rather than in heart rate. These data indicate that CO can be measured chronically and reliably in conscious mice, also in conditions of heart failure, and that variations in preload are an important determinant of CO in this species.