A game apparatus includes an operating switch and a microphone. A player operates a player object through intuition by the operating switch or inputting a sound. The number of zero crossings contained in waveform of a sound input through the microphone is detected, and also individual interval times between the zero crossings are detected. Then, it is determined whether or not the distribution of the interval times, i.e. the frequency distribution matches the distribution of interval times (frequency distribution) related to a breath sound stored in advance. If there is a match between the two, the input sound is recognized as a breath sound, and a game process based on the breath (wind) is carried out. For example, a game screen depicting the breath or wind is displayed on an LCD.