Hypoxia and acidosis are widely recognized as major contributors to the development of treatment resistant cancer. For patients with disseminated metastatic lesions, such as most women with ovarian cancer (OvCa), the progression to treatment resistant disease is almost always fatal. Numerous therapeutic approaches have been developed to eliminate treatment resistant carcinoma, including novel biologic, chemo, radiation, and photodynamic therapy (PDT) regimens. Recently, PDT using the cationic photosensitizer EtNBS was found to be highly effective against therapeutically unresponsive hypoxic and acidic OvCa cellular populations in vitro. To optimize this treatment regimen, we developed a tiered, high-content, image-based screening approach utilizing a biologically relevant OvCa 3D culture model to investigate a small library of side-chain modified EtNBS derivatives. The uptake, localization, and photocytotoxicity of these compounds on both the cellular and nodular levels were observed to be largely mediated by their respective ethyl side chain chemical alterations. In particular, EtNBS and its hydroxyl-terminated derivative (EtNBS-OH) were found to have similar pharmacological parameters, such as their nodular localization patterns and uptake kinetics. Interestingly, these two molecules were found to induce dramatically different therapeutic outcomes: EtNBS was found to be more effective in killing the hypoxic, nodule core cells with superior selectivity, while EtNBS-OH was observed to trigger widespread structural degradation of nodules. This breakdown of the tumor architecture can improve the therapeutic outcome and is known to synergistically enhance the antitumor effects of front-line chemotherapeutic regimens. These results, which would not have been predicted or observed using traditional monolayer or in vivo animal screening techniques, demonstrate the powerful capabilities of 3D in vitro screening approaches for the selection and optimization of therapeutic agents for the targeted destruction of specific cellular subpopulations.