Analysis of particle data from the CRIT II experiment, studying Alfv¿n's critical ionization velocity (CIV) effect, shows that the density of newly created ions (presumably Ba+ from the shaped-charge beam) is consistent with the increase in total plasma density measured by the independent RF plasma probe on board (Swenson et al., 1990) at the most active time period. We model this ion production using the measured electron flux data and the neutral barium model of Stenbaek-Nielsen et al. (1990a). To identify the main source mechanisms which may contribute most to the barium ionization, a simple model for barium ion density at the payload location is developed based on Liouville's theorem. We estimate that the electron impact ionization is responsible for 90% of the barium ion production observed by CRIT II in the first release and up to 45% in the second release. By employing a ''two-state approximation'' calculation (Rapp and Francis, 1962), the Ba-O+ charge exchange cross section is found to range from about 2.0¿10-17 cm2 at a velocity of 4 km/s to 2.0¿10-15 cm2 at a velocity of 20 km/s. This result suggests that the Ba-O+ charge exchange is probably dominant among all the non-CIV ionization processes. By considering the charge exchange process in our density model, the barium ion densities are calculated for the two releases on CRIT II. The comparison between the model results and the observed data is found to be reasonably consistent if the cross sections, as calculated above, are multiplied, by 0.3 for the first release and 1.0 for the second release. Our result suggests that the charge exchange process could be the most important non-CIV ionization mechanism in the CRIT II experiment and it should be considered carefully case by case in CIV experiments. ¿ American Geophysical Union 1995