A　problem　of　robust　fault　detection　for　a　class　of　networked　systems　with　uncertain　missing　measurements　probabilities　is　discussed.　The　packed-data　loss　is　modeled　as　a　Bernoulli　random　binary　switching　sequence　with　an　unknown　conditional　probability　distribution　that　is　assumed　to　be　in　an　interval.　An　observer-based　fault　detection　filter　under　all　missing　measurements　probabilities　in　the　interval　can　be　designed　to　guarantee　residual　output　sensitive　to　fault　and　robust　to　L2-norm　bounded　unknown　input.　A　sufficient　condition　is　derived　in　terms　of　linear　matrix　inequality　(LMI).　A　numerical　example　illustrates　the　effectiveness　of　the　proposed　approach.