We propose an optimal ventilation control model for complex urban vehicular tunnels with distributed pollutant discharge points. The control problem is formulated as a nonlinear integer program that aims to minimize ventilation energy cost while meeting multiple air quality control requirements inside the tunnel and at discharge points. Based on the steady-state solutions to tunnel aerodynamics equations, we propose a reduced form model for air velocities as explicit functions of ventilation decision variables and traffic density. A compact parameterization of this model helps to show that tunnel airflows can be estimated using standard linear regression techniques. The steady-state pollutant dispersion model is then incorporated for the derivation of optimal pollutant discharge control strategies. A case study of a new urban tunnel in Hangzhou, China demonstrates that the scheduling of fan operations based on the proposed optimization model can effectively achieve different air quality control objectives under varying traffic intensity.