Objective: The objective of this study is to quantify the strength of preferences and likely adherence to therapy of people with type 2 diabetes mellitus (T2DM) for outcomes of oral antidiabetic agents (OADs). Methods: Currently many of the available OADs for T2DM are associated with side effects such as weight gain and nausea and vomiting. Understanding patient preferences for medications with different attributes is important for predicting likely adherence to OAD treatments with different side-effects. We developed a web-based instrument to elicit treatment preferences. Inclusion criteria were people 18 years of age and older who were insulin-naïve and currently taking OADs to treat T2DM. The instrument included ten stated-choice trade-off tasks. Subjects chose between pairs of hypothetical medication alternatives, each including change in HbA1c levels, number of mild-to-moderate hypoglycemic events per month, water retention, weight gain in the first six months of treatment, mild stomach upset, and risk of heart attack. Sample sizes were 200 in the United States and 200 in the UK. Results: The majority of subjects were white, married, and female. Subjects were willing to make tradeoffs between all medication attributes. On average, subjects would accept significant weight gain if the medication yielded a reduction in HbA1c that resulted in optimal control (6.5%). However, subjects' stated adherence was significantly lower for treatments with weight gain, increased heart-attack risk, or persistent nausea and vomiting. The UK sample was less tolerant of hypoglycemic events. Conclusion: While people with T2DM believe glucose control is important, medication side-effects influence patients' treatment choices. Efficacy as measured under controlled, clinical-trial conditions may overstate actual treatment effectiveness as a result of patient concerns over OAD side effects.