Nicotinamide phosphoribosyltransferase (NAMPT) inhibitors (e.g., FK866) target the most active pathway of NAD(+) synthesis in tumor cells, but lack tumor-selectivity for use as a single agent. Reducing NAD(+) pools by inhibiting NAMPT primed pancreatic ductal adenocarcinoma (PDA) cells for poly(ADP ribose) polymerase (PARP1)-dependent cell death induced by the targeted cancer therapeutic, β-lapachone (β-lap, ARQ761), independent of poly(ADP ribose) (PAR) accumulation. β-Lap is bioactivated by NADPH:quinone oxidoreductase 1 (NQO1) in a futile redox cycle that consumes oxygen and generates high levels of reactive oxygen species (ROS) that cause extensive DNA damage and rapid PARP1-mediated NAD(+) consumption. Synergy with FK866+β-lap was tumor-selective, only occurring in NQO1-overexpressing cancer cells, which is noted in a majority (∼85%) of PDA cases. This treatment strategy simultaneously decreases NAD(+) synthesis while increasing NAD(+) consumption, reducing required doses and treatment times for both drugs and increasing potency. These complementary mechanisms caused profound NAD(P)(+) depletion and inhibited glycolysis, driving down adenosine triphosphate levels and preventing recovery normally observed with either agent alone. Cancer cells died through an ROS-induced, μ-calpain-mediated programmed cell death process that kills independent of caspase activation and is not driven by PAR accumulation, which we call NAD(+)-Keresis. Non-overlapping specificities of FK866 for PDA tumors that rely heavily on NAMPT-catalyzed NAD(+) synthesis and β-lap for cancer cells with elevated NQO1 levels affords high tumor-selectivity. The concept of reducing NAD(+) pools in cancer cells to sensitize them to ROS-mediated cell death by β-lap is a novel strategy with potential application for pancreatic and other types of NQO1+ solid tumors.