Treatment of chronic disorders affecting the central nervous system (CNS) is complicated by the inability of drugs to cross the blood–brain barrier (BBB). Non-viral gene therapy applied to brain capillary endothelial cells (BCECs) denotes a novel approach to overcome the restraints in this passage, as turning BCECs into recombinant protein factories by transfection could result in protein secretion further into the brain. The present study aims to investigate the possibility of transfecting primary rat brain endothelial cells (RBECs) for recombinant protein synthesis and secretion of the neuroprotective protein erythropoietin (EPO). We previously showed that 4% of RBECs with BBB properties can be transfected without disrupting the BBB integrity in vitro, but it can be questioned whether this is sufficient to enable protein secretion at therapeutic levels. The present study examined various transfection vectors, with regard to increasing the transfection efficiency without disrupting the BBB integrity. Lipofectamine 3000™ was the most potent vector compared to polyethylenimine (PEI) and Turbofect. When co-cultured with astrocytes, the genetically modified RBECs secreted recombinant EPO into the cell culture medium both luminally and abluminally, and despite lower levels of EPO reaching the abluminal chamber, the amount of recombinant EPO was sufficient to evolve a biological effect on astrocytes cultured at the abluminal side in terms of upregulated gene expression of brain-derived neurotropic factor (BDNF). In conclusion, non-viral gene therapy to RBECs leads to protein secretion and signifies a method for therapeutic proteins to target cells inside the CNS otherwise omitted due to the BBB.