Control algorithms for minimum fluidization, which identify the bed regime from information on pressure drop fluctuations, were presented for binary size mixturses. Responses to rampwise and stepwise increment of the hold up were studied experimentally among the controls adopting a critical point prediction from the operation points and the controls maintaining the aggregative condition perceiving the state from amplitude of the fluctuations. Results showed that the system was adjusted to the final condition with higher accuracy by the former algorithm than by the latter, and variance methods provided more stability compared with the method employing the amplitude measurement.