Four questions were addressed: (a) does biobehavioral intervention result in within-session reduction of tremor severity; (b) do relaxation and electromyographic (EMG) biofeedback training produce differential effects; (c) do within-session treatment effects generalize to daily performance; and (d) are reductions in tremor severity maintained at follow-up assessment? Three adults, ages 51, 77, and 83, each with a diagnosis of essential tremor (ET), and a long standing history of tremor of the hands uncontrolled by medication, took part. A repeated pre-post-training single-case experimental design embedded within a sequential A--B--C--D design was used; in addition, 1 participant received a return to the B phase. Outcome measures included within-session clinical and self-ratings of tremor severity, surface electromyography (sEMG) of forearm muscles, and daily self-ratings of tremor at home. Tremor was measured while participants engaged in eating or drinking tasks. The Behavioral Relaxation Scale (BRS) served as a process measure to assess relaxation proficiency. Clinical ratings of tremor and the BRS had high interobserver agreement. Visual inspection and statistical tests of single-case data were used to evaluate outcomes. Each participant showed significant within-session improvements on various measures of tremor and improvement during intervention as compared to baseline phases. There were no clear-cut differences between relaxation and biofeedback phases. Improvements declined somewhat at a 12-week follow-up. Relationships among measures of tremor are discussed. Biobehavioral interventions hold promise for older adults coping with ET. Further research is needed using an array of biobehavioral measures to assess intervention outcome.