The aim of the present paper is to explore in some detail the utility of parallel corpora in investigating NP modification structures in English and Hungarian. A gross statistical analysis has revealed that despite their radically different internal structure, maximally extended NP's are more suitable units for comparison than either a word class (e.g. adjective) or lexical based analysis. Among the cases when there is deviation From the A+N in either language, the paper focuses on Hungarian equivalents of English N+PP structures. The analysis leads to some useful criteria to distinguish genuine cases of explicitation form grammatical expedience.