Supporting safe self-management interventions for symptoms of osteoarthritis (OA) may reduce the personal and societal burden of this increasing health concern. Self-management interventions might be even more beneficial if symptom control were accompanied by decreased medication use, reducing cost and potential side effects. Guided imagery with relaxation (GIR) created especially for OA may be a useful self-management intervention, reducing both symptoms and medication use. A longitudinal randomized assignment experimental design was used to study the efficacy of GIR in reducing pain, improving mobility, and reducing medication use. Thirty older adults were randomly assigned to participate in the 4-month trial by using either GIR or a sham intervention, planned relaxation. Repeated-measures analysis of variance revealed that, compared with those who used the sham intervention, participants who used GIR had a significant reduction in pain from baseline to month 4 and significant improvement in mobility from baseline to month 2. Poisson technique indicated that, compared with those who used the sham intervention, participants who used GIR had a significant reduction in over-the-counter (OTC) medication use from baseline to month 4, prescribed analgesic use from baseline to month 4, and total medication (OTC, prescribed analgesic, and prescribed arthritis medication) use from baseline to month 2 and month 4. Results of this study support the efficacy of GIR in reducing symptoms, as well as in reducing medication use. Guided imagery with relaxation may be useful in the regimen of pain management for clinicians.