Due to the impairments in optical fiber transmission, regeneration is needed in certain intermediate nodes for some lengthy lightpaths. This kind of optical networks is called translucent optical networks. Spare transceivers in the optical nodes can be used for regeneration. The problem of maximizing the number of established connections (NEC) in the translucent optical networks can be formulated as a mixed-integer linear programming (MILP) problem. This formulation can only be solvable for small networks but for large networks and large traffic matrix the MILP formulation is intractable. Hence an effective heuristic algorithm based on K-least-wavelength-weight-path routing and statistical method is proposed. For small networks, which can be handled by MILP, the results show that the heuristic algorithm also yields the optimal solutions in most cases. The ability of the heuristic algorithm to handle large networks makes it suitable for use in a route engine as well as in a network design tool.