Quality　function　deployment　(QFD)　is　a　methodology　to　ensure　that　customer　requirements　(CRs)　are　deployed　through　product　planning,　part　development,　process　planning　and　production　planning.　The　first　step　to　implement　QFD　is　to　identify　CRs　and　assess　their　relative　importance　weights.　This　paper　proposes　a　nonlinear　programming　(NLP)　approach　to　assessing　the　relative　importance　weights　of　CRs,　which　allows　customers　to　express　their　preferences　on　the　relative　importance　weights　of　CRs　in　their　preferred　or　familiar　formats.　The　proposed　NLP　approach　does　not　require　any　transformation　of　preference　formats　and　thus　can　avoid　information　loss　or　information　distortion.　Its　potential　applications　in　assessing　the　relative　importance　weights　of　CRs　in　QFD　are　illustrated　with　a　numerical　example.