A new computational scheme to simultaneously optimize the electronic and ionic configurations in solids and molecules is presented in the framework of variational quantum Monte Carlo method. This scheme, in which a fictitious Lagrangian to describe the dynamics of electronic variational parameters and ionic coordinates is introduced, is formulated virtually in parallel with the Car-Parrinello method for density-functional theory. The feasibility and usefulness of the proposed scheme are demonstrated by carrying out a structural optimization for water molecule with the aid of steepest-descents technique.