More than 250,000 women will die of cancer in the United States this year, almost 10% of which are due to gynecologic malignancies. Many of these women will have received care in the intensive care unit (ICU). With important advances in medical technology and the advent of an expanded pharmacologic armamentarium, our ability to maintain life has increased greatly over the past few years. However, this phenomenon has been associated with great emotional and financial cost. It is estimated that ICU charges totaled almost 10% of the $810 billion spent on health care in the United States in 1992. Because 6-month survival rates for patients with cancer admitted to an ICU are the lowest of any disease subgroup (23.7%), we must critically evaluate the role of the ICU in the care of these patients. Decisions regarding admission to an ICU, level of care, and termination of care must take into account patient and family wishes, a reasonable estimation of the reversibility of the acute disease process in question, and the natural history of the underlying disease. Many prognostic scoring systems have been devised to estimate the probability of death among adult ICU patients; however, most of these systems were developed with data from trauma patients rather than from patients with an underlying malignancy, and none are capable of predicting which patient will die. Decisions concerning level of care in the ICU will necessarily involve medical as well as ethical considerations and are best made with a team approach.