The distributed   sampled-data filtering in sensor networks based on consensus strategy is investigated. A sampled-data filtering framework is established by transmitting sampled-data among sensors. In this framework, the filtering problem is converted equivalently into the stability of an error system with an interval time-varying delay, with the released sampled data from the sensor and its neighbor sensors as the supporting information set. A sufficient condition under which stability of the error system is guaranteed with a desired   performance is derived. Based on this condition, a distributed   filter design approach is developed. An illustrative example is given to demonstrate the effectiveness of the proposed method.