This paper comments on a critical review of cohort studies on PCB-related neurodevelopmental deficit in young children by D.V. Cicchetti, A.S. Kaufman, and S.S. Sparrow (CKS). Major points of criticism of CKS, namely alleged violation of statistical principles, presumed lack of clinical significance of findings, and alleged insufficient control of confounding are dealt with in appropriate detail. It is argued that much of this criticism is inappropriate and biased, and that, in particular, arguments dealing with basic rules of statistical inference rely too heavily on the narrow principles of Neyman-Pearson testing which are discussed controversially in modern epidemiology. Other critical arguments concerning the presumed lack of validity of assessment instruments, the apparent lack of reliability checks in some studies, and the presumed inappropriate treatment of longitudinal data are also discussed as being poorly founded. It is finally concluded that, although there are inconsistencies and weaknesses both within and between individual PCB-studies, the almost unconditional rejection of the full set of cohort studies by CKS is in no way supported by good reasoning.