This paper aims to account for the idiosyncratic properties of the Contrastive Stripping Construction (CSC) in English. The main issues on this construction are how to generate its 4 possible skeletal structures and get the right contrastive reading from them. Rejecting the Movement-Ellipsis approaches facing theoretical and/or empirical problems, we propose a constructionbased analysis of the CSC which interacts the English CSC-cx rule with various lexical information of the lexemes such as and and but. This enables us to generate the 4 skeletal structures of the CSC and assign an appropriate contrastive reading to them.