The long term goal of this research program is to relieve the unnecessary associated with cancer pain by effectively combining pharmacological and behavioral interventions. Over a million Americans are diagnosed with cancer annually; 85% of these people will experience moderate to severe pain. Although pharmacological therapies do exist to manage the pain for 95% of these people, significant barriers often impede management and 25% die with unrelieved pain. Inaccurate pain inferences by clinicians are a major barrier. Clinicians infer more pain when patients exhibit expressive behaviors. Most patients with cancer, however, do not exhibit pain with expressive behavior. They often appear stoic and try not to communicate their pain to other people, but stoicism may hinder communication with clinicians who are responsible for prescribing pain therapies. Furthermore, patients perform cancer pain control behaviors which may be misinterpreted by clinicians as indications of a pain-free state. In this project, 3 studies will be conducted to: 1) test a method of coaching patients to inform clinicians about their pain, 2) provide further evidence for the pain control behavior construct, and 3) pilot test extension of the coaching method in patients with lung cancer. 1)In 200 patients with lung cancer, a randomized clinical trial of a COACHING protocol will be tested for effects on clinicians' prescribing behavior and patients' pain, coping behavior and emotional status. It is hypothesized that patients randomly assigned to receive 4 COACHING sessions will verbally communicate increased amounts of pain data and have decrease pain intensity, pain coping attempts, anxiety, and depression 6 weeks after the first COACHING session. Additionally, clinicians' inferences about pain will improve for the COACHED patients. 2) A descriptive study will provide further description of lung cancer pain control behavioral observation of patients engaging in standardized activities of daily living will be used in the 200 patients participating in the clinical trial. 3)Results of the two studies will be used to develop and pilot test an extension of the coaching protocol that will assist patient to self-monitor sensory components of the cancer pain experience and self-manage the pain by using optimal pain control behaviors. Accurate inference is critical for clinicians to change analgesic therapy plans when pain is not adequately controlled. COACHING has the potential t dramatically improve cancer pain management and patients' quality of life.