Nanobubbles (NBs) have attracted increasing attention due to their unique physicochemical characteristics and enormous potential industrial applications in the biology, environment, medicine, and many other fields. A complete understanding of the underlying stability mechanism of NBs is an essential requirement for their research, development, and industrial applications. In this study, a facile and reliable approach to generating different gas NBs using a porous ceramic membrane is reported. In addition, an improved method of statistical analysis to predict NB size distribution is proposed. The results indicated that the gas NBs presented a log-normal distribution due to an underlying Ostwald ripening process. The concentration of NBs in solution was controlled by the gas type owing to different magnitude of molecular polarizability of dissolved gases. The mean size of the NB was inversely proportional to the surface charge density. Moreover, the NB stability lies in a special state at the bubble interface, not only because the presence of surface charges hinders collisions between bubbles, but also because the degree of gas supersaturation of molecules around bubbles resulting in gas exchange has a certain stabilizing effect, and also controls the evolution of particle size distribution (PSD) for different bulk NBs.