In the past, people with an incurable cancer were not informed of the reality of their condition—a charade of “don’t tell/ don’t ask” was played by physician and patient. Now, we work under the assumption that patients and their families deserve the truth. How can hope be sustained without elision or fabrication? Helft believes that this is possible through “collusion.” He builds an argument that it is ethical, indeed beneficial, to communicate with patients in a metered way, gradually adding or subtracting a degree of hope based on unfolding clinical events. This collusive approach is rooted in the paradox of uncertainty. Everyone with illness, but most squarely those with severe and lifethreatening disease, faces the fact that nothing in medicine is absolute. On the one hand, this is a source of anxiety, because there are no guarantees that a drug or procedure will succeed one hundred percent of the time. On the other hand, the absence of absolute certainty provides a basis for hope. Even the worst disorders have the potential for an unexpected outcome. Yes, their outcomes statistically cluster around a median and a mean, but there also can be cases at the far end of the distribution. After years in practice, most oncologists encounter these extended limits of the curve. There is the unusual case of a patient with advanced malignancy whose remission lasted years instead of months, or the even rarer case who completely defied the odds and entered a sustained complete remission. I still care for a woman, who more than 20 years ago, presented with inflammatory carcinoma of the breast and an underlying 8 cm primary mass; her tumor has waxed and waned with no significant clinical morbidity on treatment with a series of hormonal agents. A physician colleague of mine with metastatic gastric carcinoma has no evidence of disease 17 years after surgery, chemotherapy, and radiation. One woman with multiple cerebral metastases from breast cancer had 5 years of quality of life before relapsing and dying. And patients I saw as a consultant later died of “natural causes,” following many years in remission from CNS lymphoma or metastatic non–small-cell carcinoma of the lung. Some are critical of doctors who refer to such rare cases and the uncertainty that they illustrate. These critics assert that rarities are a wellspring of false hope for people who should squarely face their imminent demise. They further contend that raising the specter of a positive outcome manipulates patients and their families into pursuing therapies that are toxic and clinically worthless, therapies that cost society large sums. The strategy of collusion largely deflects such criticisms, since the approach is sculpted to the clinical course of the individual. If the initial period of treatment proves terribly toxic and ineffective against the malignancy, then this result serves as a platform for reassessment of the therapeutic plan and focuses the patient and his family on the more certain negative outcome. But if the tumor regresses and the patient can enjoy some quality of life with the particular treatment, then it becomes more probable that the person’s own biology and the biology of the tumor may move him further along the outcome curve, away from the median and mean. Helft convincingly argues it is From the Harvard Medical School, Division of Experimental Medicine, Beth Israel Deaconess Medical Center, Boston, MA.