OBJECTIVES:  To elicit the preferences of myeloma patients for different outcomes in the treatment of advanced cancer. METHODS:  Treatment benefits were quantified in terms of probability of survival for two years or longer. Risks were quantified in terms of probability of experiencing mild to moderate side-effects for two months or longer and the probability of experiencing severe to life-threatening side-effects lasting up to two months. Ordinal statements expressing the relative importance of a change on one criterion compared to a change on another (e.g., increasing the probability of survival from 50% to 60% is more important than decreasing the probability of experiencing severe side-effects from 80% to 50%) were collected through an online survey. For each patient, an underlying, representative value function was subsequently estimated by first repeatedly sampling from the space of all additive value functions consistent with the provided preference statements and then taking the average of the sampled functions. RESULTS:  A total of 563 patients (median age between 61 and 70, 56% male) completed the online questionnaire. The majority of these patients were currently either in remission following treatment (53%) or receiving treatment (39%) and the median time since diagnosis of myeloma was between 4 and 6 year. Most patients considered increasing the survival probability more important than decreasing the probability of experiencing the two types of side-effects, but the weights attached to each of these criteria varied greatly across patients. CONCLUSIONS:  Quantitative studies conducted in a larger group of patients provide useful information on the variability between individual preferences and may be helpful in identifying subgroups of patients with homogeneous preferences.