Presents analytic criteria for the local activity theory in two-port cellular neural network (CNN) cells with four local state variables, and gives the application to a smoothed Chuas circuit (SCC) CNN with two-port and 1515 arrays. The bifurcation diagrams of the SCC CNN show that they are completely the same as those of an SCC CNN with one-port calculated earlier, which do not exist locally passive domain. The evolution of the patterns of the state variables of the SCC CNN is stimulated. Oscillatory patterns, chaotic patterns, or divergent patterns may emerge if the selected cell parameters are located in the locally active unstable domains but nearby the edge of chaos domain.