In　this　paper,　we　propose　a　variable　selection　procedure　for　partially　linear　varying　coefficient　model　under　quantile　loss　function　with　adaptive　Lasso　penalty.　The　functional　coefficients　are　estimated　by　B-spline　approximations.　The　proposed　procedure　simultaneously　selects　significant　variables　and　estimates　unknown　parameters.　The　major　advantage　of　the　proposed　procedures　over　the　existing　ones　is　easy　to　implement　using　existing　software,　and　it　requires　no　specification　of　the　error　distributions.　Under　the　regularity　conditions,　we　show　that　the　proposed　procedure　can　be　as　efficient　as　the　Oracle　estimator,　and　derive　the　optimal　convergence　rate　of　the　functional　coefficients.　A　simulation　study　and　a　real　data　application　are　undertaken　to　assess　the　finite　sample　performance　of　the　proposed　variable　selection　procedure.