The paper discusses centralised integration of multisensor noisy and fuzzy data, which employs both Kalman filtering and fuzzy arithmetic. Because of the property of fuzzy arithmetic, the parameter fuzziness in a system under extended operation will increase without limit and finally reach an unacceptable range. Hong and Wang adopted a new compression technique to solve this problem. The paper extends their work on single sensor noisy and fuzzy data filtering to multisensor noisy and fuzzy data filtering. An example is given to illustrate the effectiveness of the algorithm presented.