Vagal nerve stimulation (VNS) has been explored as a potential therapy for chronic heart failure. The contribution of the afferent pathway to myocardial interstitial acetylcholine (ACh) release during VNS has yet to be clarified. In seven anesthetized Wistar-Kyoto rats, we implanted microdialysis probes in the left ventricular free wall and measured the myocardial interstitial ACh release during right VNS with the following combinations of stimulation frequency (F in Hz) and voltage readout (V in volts): F0V0 (no stimulation), F5V3, F20V3, F5V10, and F20V10. F5V3 did not affect the ACh level. F20V3, F5V10, and F20V10 increased the ACh level to 2.83 ± 0.47 (P < 0.01), 4.31 ± 1.09 (P < 0.001), and 4.33 ± 0.82 nM (P < 0.001), respectively, compared with F0V0 (1.76 ± 0.22 nM). After right vagal afferent transection (rVAX), F20V3 and F20V10 increased the ACh level to 2.90 ± 0.53 (P < 0.001) and 3.48 ± 0.63 nM (P < 0.001), respectively, compared with F0V0 (1.61 ± 0.19 nM) but F5V10 did not (2.11 ± 0.24 nM). The ratio of the ACh levels after rVAX relative to before was significantly less than 100% in F5V10 (59.4 ± 8.7%) but not in F20V3 (102.0 ± 8.7%). These results suggest that high-frequency and low-voltage stimulation (F20V3) evoked the ACh release mainly via direct activation of the vagal efferent pathway. By contrast, low-frequency and high-voltage stimulation (F5V10) evoked the ACh release in a manner dependent on the vagal afferent pathway.