Ovarian cancer is a leading cause of cancer deaths in women. Only a portion of women diagnosed with this cancer will respond to conventional drugs, and alternative therapies only work in some cases. In order to better tailor treatment, personalized medicine has turned to ‘markers’ that statistically evaluate the chances of a drug being effective. Here we propose a completely different approach, where we will directly test in the tumor cells themselves their ability to respond to multiple drugs. While this approach has been tried before, it has largely failed because it didn’t take into account the specific features of the tumor tissue, wasn’t economical enough, and couldn’t be performed sufficiently rapidly to influence clinical decision-making. By using a miniaturized system, coupled to the novel model systems we have developed over the last several years, we propose to rapidly test either complex three dimensional cell structures, or even miniscule pieces of tumor tissue, for their response to therapy. The fact that we can in parallel test different samplings of any given tumor, as well a rapid and real-time readout of whether the tumor cells are dying, will provide us with a way we can efficiently and cost-effectively determine whether a specific treatment will be of benefit to the patient, thus improving outcomes and managing health costs.