Use of an ultrasound contrast agent in echocardiography increases the probability for generation of premature ventricular contractions (PVCs), and the probability is particularly high when image acquisition is triggered at the end-systolic phase. In this study, the relation between timing of ultrasound exposure and probability of PVC generation was investigated using cultured cardiac myocytes. Myocytes that had been isolated from neonatal rats and cultured on a cover slip were used in this study. These myocytes start autonomous pulsation after incubation for several days. Continuous monitoring of the pulsation was done by detecting changes in the light intensity of the microscopic image of a myocyte. Pulsed ultrasound with peak-negative pressures of 0.27 MPa and 1.1 MPa was used in the presence and absence of microbubbles, respectively. Generation of PVCs was examined at various delay timings of ultrasound exposure after the onset of myocyte contraction. In cases both with and without microbubbles, PVCs were generated only when myocytes were exposed to ultrasound with delay times above a certain threshold of time. The results showed that each myocyte has a different threshold for producing PVCs, and the delays ranged from 74 to 310 ms; however, the standard deviation of these thresholds observed in each myocyte was small, an average of 8.2 ms. Averaged thresholds in the groups of cells exposed to ultrasound with and without bubbles were 193.6 and 210.3 ms, respectively, suggesting that the presence of bubbles does not change the mechanism by which PVCs are induced.