The primary aim of our thesis is to develop a modal query language. We want to select fragments of modal logics that are solvable in polynomial time, and then formulate them into a query language form. We also want to develop a bottom-up computation method for the selected fragments. Another problem we want to study is whether decision procedures and upper space bounds for propositional modal logics can be improved.