Background: Major depressive disorder is associated with elevated levels of DNA oxidation, DNA damage, and gene expression of DNA repair enzymes including poly (ADP-ribose) polymerase-1 (PARP1). Elevated PARP1 activity is directly linked to neuroinflammation and PARP inhibitors are anti-inflammatory and neuroprotective. We previously showed that PARP inhibitors produce antidepressant-like effects equivalent to fluoxetine in rodent models. Here, we examined whether the PARP inhibitor 3-aminobenzamide (3AB) is effective in a rat model of treatment-resistant depression. Methods: Treatment-resistant depression was modeled with injections of lipopolysaccharide (LPS; 0.1 ug/kg/day) and daily chronic unpredictable stress (CUS) for 28 days. Anhedonia and helplessness were indexed with sucrose preference and forced swim tests, respectively, in 5 groups of rats (n¼6-8 rats/group) including unstressed, CUS, and CUS+LPS rats treated with saline, and CUS+LPS rats treated with either 3AB or fluoxetine. Results: Anhedonia induced by CUS+LPS was significantly attenuated by 3AB (p¼0.01), while fluoxetine failed to do so. Likewise, 3AB was superior to fluoxetine in reducing helplessness, where latency to immobility times were significantly lower in CUS+LPS rats treated with fluoxetine (p¼0.001) compared to unstressed rats, but not significantly different for 3AB-treated CUS+LPS rats. Conclusions: The PARP inhibitor 3AB demonstrated robust and unique antidepressant activity superior to fluoxetine in the TRD rat model. PARP is linked to neuroinflammation through release of microglia-activating factors including poly (ADP-ribose) and HMGB1, and through NF-kB activation, pathways under investigation by our lab. PARP inhibitors are currently used clinically to facilitate cytotoxicity of DNA-damaging anti-cancer treatments. Further research could implicate re-purposing non-cytotoxic PARP inhibitors for treatment-resistant depression.