It is shown that the quantitative interpretation of recent experiments to determine the ion drag force in complex ͑dusty͒ plasmas ͓C. Zafiu, A. Melzer, and A. Piel, Phys. Plasmas 9, 4794 ͑2002͒; 10, 1278 ͑2003͔͒ is not correct. A comparison of different models of the ion drag force is carried out to illustrate the complexity of this issue and to highlight the current level of the research.There have been many experiments, both on the ground 1-3 and under microgravity conditions, 4 -7 which demonstrate that dust particles in rf plasmas usually do not fill the entire plasma volume, but instead a particle-free region-a void-is formed in the central part of the discharge. Such behavior was attributed by Samsonov and Goree 8 to the action of the ion drag force which points outwards and can exceed the electric force which pushes particles to the center. If the ion drag is stronger than the electric force in the center of a discharge ͑where electric fields are weak͒, then the void can be formed. This mechanism can explain all qualitative features observed so far in experiments with the void. This is why the ion drag is generally considered the most prominent candidate for explaining the particle repulsion from the central region of rf discharges. 9,10 There were, however, considerable problems with the quantitative description of the void formation: The value of the ion drag force calculated from the standard model by Barnes et al. 11 was systematically smaller ͑by about an order of magnitude͒ than the electric force. 4,12 Nevertheless, in recent papers Zafiu et al. 13,14 claim that the ion drag force calculated from the Barnes model is quite sufficient to explain the void formation-in contrast to all the previous works above. It is the purpose of this Comment to resolve this inconsistency.The root of the discrepancy lies in a single assumption. The model by Barnes et al. 11 assumes that '' . . . no ion interaction with the particle occurs outside of a Debye length.'' Therefore, the magnitude of the ion drag force strongly depends on the value of the Debye ͑screening͒ length. Zafiu et al. 14 postulate the plasma screening length to be equal to the electron Debye length. This choice of the screening length can be appropriate for rf sheaths or dc striations, where the electric fields are very strong. Then the ion drift is supersonic and hence ions cannot participate in the screening of particles. 15,16 But this is certainly not the case in the central region of an rf discharge, where the ͑ambipolar͒ electric fields are weak ͑the field equals zero at a point of the maximum plasma potential͒ and the ion drift is usually ͑sub͒ther-mal. In experiments 13,14 the measurements were performed close to the center of an rf chamber, and the upper estimate of the ratio of the ion drift velocity to the thermal velocity does not exceed Ӎ2 for the lowest pressure. 14 For such conditions the relevant plasma screening length is the linearized Debye length D ϭ( Di Ϫ2 ϩ De Ϫ2 ) Ϫ1/2 , which is close to the ion Debye length Di ͑typically a...