A mathematical model of active neuronal dendrite is constructed in the discrete form based on physiological informations. The dynandcs of this model is described by calcium-activated regenerative processes with threshold and calcium accumulation system which changes the dynamical properties of the model. Our model is capable of reproducing a stable oscillation in response to a maintained current stimulation. The firing rate in response to the periodic stimulation monotonically increases in a stepwise manner as a function of the amplitude of stimuli. This stimulus-response properties are commonly observed in aM actual neurons and neuron models. The model can also reproduce caldumactivated plateau potential based on bifurcation mechanism parametrized by an intracellular calcium concentration. These dynamical properties coincide with that of actual neuron. Compaitmental dendrite model is constructed by coupling these single compartment models to make up dendritic structure. Active potentials propagating on compaitmental dendrite model mutually interact so that the propagating velocity is modified. Active dendrite which is constructed with the "neuron" could process information throufh its complicated dynamics.