A single cell growth image system was used to study the growth of Pseudomonas aeruginosa single cells. A stochastic modeling process was applied to connect the growth of P. aeruginosa single cells and cell populations, which made it possible to simulate the growth of P. aeruginosa . The lag time and specific growth rate distributions with different inoculum sizes were obtained as a result of the simulation’s repetitive executions which were further introduced into the reduced Baranyi model for establishing an individual-based model. Then a stochastic growth process of P. aeruginosa was conducted by using Monte Carlo simulation. Results showed that a negative relationship existed between lag time and inoculum size. As the inoculum size increased from 1 cell to 100 cells, the lag time decreased from 2.91 h to 2.55 h at 25℃ and from 1.49 h to 0.99 h at 35℃. The coefficient of variation decreased from 29.90% to 2.96% at 25℃ and from 22.53% to 4.64% at 35℃. The specific growth rate was more affective to the temperature changes which increased from 0.70 lnCFU/h at 25℃ to 1.00 lnCFU/h at 35℃. Meanwhile, the stochastic growth of P. aeruginosa with different inoculum sizes demonstrated that the growth of P. aeruginosa showed a determinate tendency as inoculum sizes increasing, in spite of the stochastic growth property of bacterial single cells. Compared with the traditional determinate predictive modelling, studying bacterial population growth from stochastic single cell dynamics opened the door for applications in risk assessment and prediction of shelf life.