We recently described a non-viral gene therapy paradigm offering long-term resolution of established neuropathic pain in several animal models. Here, the requirements for long-term therapeutic effects are described, and evidence is provided for a mechanism of action based on immunological priming of the intrathecal (i.t.) space. Long-term pain reversal was achieved when two i.t. injections of various naked plasmid DNA doses were separated by 5h to 3 days. We show that an initial DNA injection, regardless of whether a transgene is included, leads to an accumulation of phagocytic innate immune cells. This accumulation coincides with the time in which subsequent DNA injection efficacy is potentiated. We show the ability of non-coding DNA to induce short-term pain reversal that is dependent on endogenous interleukin-10 (IL-10) signaling. Long-term efficacy requires the inclusion of an IL-10F129S transgene in the second injection. Blockade of IL-10, by a neutralizing antibody, either between the two injections or after the second injection induces therapeutic failure. These results show that this gene therapy paradigm uses an initial 'priming' injection of DNA to induce accumulation of phagocytic immune cells, allowing for potentiated efficacy of a subsequent 'therapeutic' DNA injection in a time- and dose-dependent manner.