Despite the provision of antiemetic agents in accordance with published American Society of Clinical Oncology (ASCO) guidelines, chemotherapy-related nausea remains a clinically significant issue that is rated by patients as a greater problem than chemotherapy-related vomiting. Nausea following treatment is three times more likely to occur than vomiting. Primary Aim 1 of this study examines whether control of nausea in patients who experienced chemotherapy-induced nausea and vomiting (CINV) following their initial chemotherapy despite receiving the appropriate ASCO-recommended antiemetics can be improved by the addition of either prochlorperazine (Compazine®) or olanzapine (Zyprexa®) on days 1-4. Current ASCO guidelines for refractory CINV suggest that oncologists consider adding olanzapine (Zyprexa®) or a dopamine antagonist such as prochlorperazine (Compazine®) to the antiemetic regimen, but these are only two of many possible strategies for control of refractory CINV, none of which has been empirically tested. Primary Aim 2 will test whether olanzapine, which is a newer, more expensive antiemetic drug than prochlorperazine, is more effective than prochlorperazine in controlling nausea when used in combination with aprepitant, palonosetron and dexamethasone. This study follows up on the most recent of our five multicenter CINV studies. We will also address an additional problem regarding control of chemotherapy-related nausea. That is the lack of empirically-based models predicting chemotherapy-induced nausea from common moderately or highly emetogenic chemotherapeutic agents that take into account not only receipt of a state-of-the-art antiemetic regimen but also patient factors such as age, race, ethnicity, alcohol consumption, expectancy, anxiety, degree of nausea on the morning prior to treatment, and prior history of nausea. This prediction model will be an important addition to the antiemetic guidelines. The proposed study consists of two parts with screening and some assessments occurring at Cycle 1 and the randomized portion of the study (N = 334) occurring at Cycle 2. At Cycle 1, we will consent chemotherapy naïve breast cancer patients about to begin one of four specified chemotherapy regimens with high/moderate emetogenic potential and scheduled to receive an antiemetic regimen that conforms to ASCO Clinical Practice Guidelines. We anticipate needing to consent approximately 800 patients at Cycle 1 to meet our Cycle 2 target number. The Cycle 2 portion will be conducted in those patients who experienced moderate or greater nausea at Cycle 1. It will be a Phase III randomized, double-blinded, placebo-controlled, 2-arm study (N = 334) that builds upon our prior CINV studies and investigates optimal control of CINV in patients who experienced CINV following initial chemotherapy. This study will be implemented by the University of Rochester Cancer Center (URCC) NCI Community Oncology Research Program (NCORP) Research Base that our office manages.