A　fuzzy　non-linear　programming　model　is　developed　for　the　assessment　of　filter　allocation　and　replacement　strategies　in　fluid　power　systems　(FPSs)　under　uncertainty.　It　can　not　only　handle　uncertainties　expressed　as　L-R　fuzzy　numbers,　but　also　enhance　the　system　robustness　by　transforming　the　fuzzy　inequalities　into　inclusive　constraints.　In　modelling　formulation,　theory　of　contamination　wear　is　introduced　to　reflect　the　interactions　between　system　performance　and　system　contamination　level.　The　developed　method　has　been　applied　to　the　planning　of　filter　allocation　and　replacement　for　an　FPS　with　a　bypass　filtration　system,　and　a　piecewise　linearization　approach　is　proposed　for　solving　model　non-linearities.　Three　different　contaminant　ingression　rates　are　examined　based　on　a　number　of　filter-installation　scenarios.　The　generated　solutions　can　be　used　for　providing　guidance　to　decision　makers　to　identify　better　contamination　control　plans　for　achieving　a　minimized　system　cost　and　system-failure　risk.