OBJECTIVES: To determine whether treating conditions having medically unexplained symptoms is associated with lower physician satisfaction and higher ascribed patient responsibility, and to determine whether higher ascribed patient responsibility is associated with lower physician satisfaction in treating a given condition. METHODS: We surveyed a nationally representative sample of 1504 US primary care physicians. Respondents were asked how responsible patients are for two conditions with more-developed medical explanations (depression and anxiety) and two conditions with less-developed medical explanations (chronic back pain and fibromyalgia), and how much satisfaction they experienced in treating each condition. We used Wald tests to compare mean satisfaction and ascribed patient responsibility between medically explained conditions and medically unexplained conditions. We conducted single-level and multilevel ordinal logistic models to test the relation between ascribed patient responsibility and physician satisfaction. RESULTS: Treating medically unexplained conditions elicited less satisfaction than treating medically explained conditions (Wald P < 0.001). Physicians attribute significantly more patient responsibility to the former (Wald P < 0.005), although the magnitude of the difference is small. Across all four conditions, physicians reported experiencing less satisfaction when treating symptoms that result from choices for which patients are responsible (multilevel odds ratio 0.57, P = 0.000). CONCLUSIONS: Physicians experience less satisfaction in treating conditions characterized by medically unexplained conditions and in treating conditions for which they believe the patient is responsible.