A rich literature on Icelandic syntax has established that infinitival complements of obligatory control verbs constitute a case assignment domain independent from the matrix clause, and in this differ systematically from all types of A-movement, which manifest case dependence/preservation. As Landau (2003) has observed, these facts provide significant counterevidence to the movement theory of control (Hornstein 1999 and subsequent work). Boeckx and Hornstein (2006a) attempt to defend this theory in light of data from Icelandic. We offer here a review of the relevant literature, and we show that Boeckx and Hornstein's reply fails on several counts. We further argue that contrary to their claims, PRO in Icelandic receives structural rather than default (nominative) case, leaving the movement theory with no account for the distinction between PRO and lexical subjects.