Chomsky (2000) proposes that the operation Attract-F of Chomsky (1995) be replaced by the operation Agree, in which there is, in principle, no movement at all. In the Agree theory, agreement between a Probe P and its Goal G can lead to feature checking in situ. Lasnik (2002), however, argues that feature movement (or attract) is to be preferred to agreement at a distance, presenting three pieces of evidence: pseudo-gapping, sluicing, and remnant movement. The purpose of this article is to show that the agreement-at-a-distance theory of Chomsky (2000) can still be maintained if we adopt Pesetsky and Torrego`s (2001) conception of the EPP property as a "subfeature of a feature." Furthermore, if we assume Boeckx and Stjepanovic´`s (2001) claim that head movement is a PF phenomenon, the persuasiveness of Lasnik`s (2002) argument becomes even weaker. Following Chomsky`s (2001b) assumption that internal Merge can apply either before or after Spell-Out, I have demonstrated that even the remnant movement phenomenon, the most powerful piece of evidence for Lasnik`s (2002) argument, can naturally be accounted for with the feature checking in situ of the Agree theory.