In this paper, we present an optimal filter for linear time-invariant continuous-time stochastic systems that simultaneously estimates the states and unknown inputs in an unbiased minimum-variance sense. The optimality of the proposed filter is proven by reduction to an equivalent system without unknown inputs. Then, a second proof is given for a special case by limiting case approximations of the optimal discrete-time filter , thus establishing the connection between the continuous- and discrete-time filters. Conditions for the existence of a steady-state solution for the proposed filter are also given. Moreover, we show that a principle of separation of estimation and control holds for linear systems with unknown inputs. An example is given to demonstrate these claims.