Designing solid-state perovskite oxide solar cells with large short circuit current ( J SC ) and open circuit voltage ( V OC ) has been a challenging problem. Epitaxial BiFeO 3 (BFO) films are known to exhibit large V OC (>50 V). However, they exhibit low J SC (≪μA/cm 2 ) under 1 Sun illumination. In this work, taking polycrystalline BiFeO 3 thin films, we demonstrate that oxygen vacancies (V O ) present within the lattice and at grain boundary (GB) can explicitly be controlled to achieve high J SC and V OC simultaneously. While aliovalent substitution (Ca 2+ at Bi 3+ site) is used to control the lattice V O , Ca and Ti cosubstitution is used to bring out only GB-V O . Fluorine-doped tin oxide (FTO)/Bi 1- x Ca x Fe 1- y Ti y O 3-δ /Au devices are tested for photovoltaic characteristics. Introducing V O increases the photocurrent by four orders ( J SC ∼ 3 mA/cm 2 ). On the contrary, V OC is found to be <0.5 V, as against 0.5-3 V observed for the pristine BiFeO 3 . Ca and Ti cosubstitution facilitate the formation of smaller crystallites, which in turn increase the GB area and thereby the GB-V O . This creates defect bands occupying the bulk band gap, as inferred from the diffused reflection spectra and band structure calculations, leading to a three-order increase in J SC . The cosubstitution, following a charge compensation mechanism, decreases the lattice V O concentration significantly to retain the ferroelectric nature with enhanced polarization. It helps to achieve V OC (3-8 V) much larger than that of BiFeO 3 (0.5-3 V). It is noteworthy that as Ca substitution maintains moderate crystallite size, the lattice V O concentration dominates GB-V O concentration. Notwithstanding, both lattice and GB-V O contribute to the increase in J SC ; the former weakens ferroelectricity, and as a consequence, undesirably, V OC is lowered well below 0.5 V. Using optimum J SC and V OC , we demonstrate that the efficiency ∼0.22% can be achieved in solid-state BFO solar cells under AM 1.5 one Sun illumination.