This paper proposes a probabilistic optimiz-ation strategy for optimized use of Photovoltaic (PV) generators in distribution networks voltage regulation. Probabilistic PV output scenarios are incorporated in a constrained optimal load flow planning analysis for a given distribution network to minimize a voltage deviation index indicative of the quality of voltage regulation. The proposed strategy utilizes the PV generators as capacitors and inductors, whose reactive power compensation is monthly scheduled in coordination with the high-voltage/medium-voltage (HV/MV) substation on-load tap changer setting. The proposed strategy does not rely on a communication infrastructure nor does it comprise feedback voltage control. Physical solar irradiance and ambient temperature measurements are used to demonstrate the effectiveness of the proposed strategy on two detailed North-American MV distribution networks (25 kV and 12.66 kV). The results are validated by two distribution load flow calculation algorithms: backward/forward sweep in a MATLAB script; and the harmonic load flow solver in OpenDSS.