This study aimed to assess the long-term outcome of progressive exercise and local pulsed ultrasound in the treatment of 30 chronic tennis elbow patients (2 men, 18 women, mean age 42.3 years). The patients were originally randomised into 1) four-step progressive exercise (EX, n = 16) and 2) local pulsed ultrasound (US, n = 14) treatment groups. Before the beginning of the treatment, the groups were similar in terms of pain scores, sick-leave days and duration of symptoms. The patients underwent an 8-week treatment intervention. Long-term follow-up evaluation of the patients was performed 1) prospectively using a pain questionnaire on VAS and pain drawings classified into 5 categories, and 2) retrospectively with a postal questionnaire (which was sent to the patients to fill in. Sick-leave days, medical and physiotherapy visits, operations, early retirements and job relocations were inquired in the postal questionnaire. The diagnosis-related sick-leave days of the patients were collected from the Database of the Social Insurance Institution of Finland and the number of operations from the local hospital register. Twenty-three patients (12 in the EX group and 11 in the US group) responded. The mean follow-up time was 36 months. After the treatment the patients in the EX group needed significantly less physiotherapy (p = 0.02), fewer medical consultations (p = 0.005) and other treatments and had fewer sick-leave days (p = 0.005) than before the treatment intervention. The patients in the US group had after the treatment intervention more 17 medical visits (ns), 291 sick-leave days (ns) and less 95 physiotherapy visits (ns) than before the treatment. Eight patients (67%) in the EX group and 5 (45%) in the US group still held their previous job, while two patients in the US group, but none in the EX group were absent from work because of the tennis elbow syndrome. The patients in the EX group reported significantly lower pain scores on VAS than those in the US group. The mean pain drawing category was 1.5 in the EX group and 2.7 in the US group (p = 0.008). All the pain scores and pain drawing categories in the EX group had changed to be significantly better than in the US group, where only pain under strain had significantly improved. Because of resistant symptoms, 5 patients were operated in the US group and one in the EX group. Neither spontaneous healing, nor self-limiting of the disorder were noted during the follow-up period. The progressive exercise evaluated in this study showed beneficial long-term effects compared to ultrasound treatment in terms of pain alleviation and working ability, and the functional overall condition of the exercise patients was also better. Exercise may be able to prevent chronicity and should hence be tried and recommended.