Uncertainty in the kinetics of the second drying period represents the major challenge in continuous drying of particulate solids which can be overcome by adaptive feedback control. In this contribution, self-tuning control comprising online-state estimation and adaptive PI-control is considered to guarantee desired product moisture. Advantages of the approach over conventional PI-control are discussed. The performance is shown for fluidized bed drying of baker's yeast. It is shown that the self-tuning approach is superior to nonadaptive approaches in case of parametric uncertainty even at non-nominal steady-states.