This paper argues that the syntactic derivations underlying intensional transitive want and need are distinct. Recent proposals have attempted to provide a unified syntax for these predicates (e.g., Larson et al. 1997). Most syntactic analyses of want and need argue in favor of a derivation that involves a silent embedded verb HAVE. However, cross-linguistic evidence suggests that such analyses cannot be correct. It is shown that while only Indo-European HAVE-languages have transitive need, both HAVE and BE-languages have transitive want, suggesting that HAVE cannot be a core component in the derivation of want cross-linguistically, although it may be for need.