Twenty-five percent of older adults experience significant osteoarthritis (OA)-related comorbid sleep disturbance. Insomnia is associated with substantial negative effects on function, mood, and medical resource utilization. Cognitive behavioral therapy for insomnia (CBT-I) is evidence- based and has been shown to be efficacious in populations with a variety of comorbid conditions including OA-related chronic pain. However, in-person CBT interventions are unlikely to be widely deployable in healthcare systems. Telephone delivery has the advantage of giving patients access to personalized, efficacious CBT-I interventions from home, increasing generalizability, and outreach to minority, rural, and other underserved populations. Previous small trials of telephone CBT-I have reported positive insomnia outcomes, but no large randomized trials have assessed its long-term efficacy, impact on related outcomes in a primary care patient population, or cost-effectiveness. This proposal builds upon our successful completion of two randomized trials of CBT for insomnia with older adults: Lifestyles, a trial of group CBT in older adults with OA-related insomnia and pain, and MsFLASH04, a trial of telephone CBT-I for insomnia in menopausal women with comorbid vasomotor symptoms. Older (60+ yrs) primary care patients across Washington State will be screened for severe persistent OA-related insomnia and pain. Two hundred and seventy patients will be randomized to either CBT-I or an education only attention control (EOC). Each treatment will consist of six 20-30 minute telephone-based sessions over an eight week period. Pre-treatment, post-treatment (2 months), and 9 and 18 month assessments will include measures of sleep, pain, fatigue, mood, and quality of life. A cost effectiveness evaluation of the intervention will also be conducted. We hypothesize: 1) CBT-I will produce significantly greater initial and long-term improvements in sleep outcomes relative to EOC; 2) CBT-I will produce significantly greater initial and long-term improvements in pain, fatigue and mood relative to EOC; and 3) CBT-I will produce long-term reductions in health care utilization and costs relative to EOC. We will also explore whether changes in insomnia severity explain observed effects of CBT-I on secondary outcomes (mediator analysis), and whether CBT-I effects on insomnia differ by baseline insomnia, pain, or depression symptom severity (moderator analysis). The proposed research will determine if telephone CBT-I improves OA insomnia and associated outcomes in a state-wide primary care population of older adults, and inform policy decisions about widespread dissemination of telephone CBT-I in this and related patient populations.