For Markov jumping linear systems (MJLS), it is often encountered that the jumping mode cannot be accessible for filtering. Therefore it is strongly desirable to design a filter which is independent of the jumping model. In this paper, a new deterministic filter design procedure is proposed, which shows less conservatism than existing results. The whole design procedure can be accomplished by solving a set of linear matrix inequalities (LMIs). Finally, numerical examples are given to illustrate the effectiveness of the proposed approach. (C) 2008 Elsevier Ltd. All rights reserved.