Objective: The objective of this work was to assess the effect of physical therapy in patients with somatosensory tinnitus (ST) and explore the influence of physical therapy on clinical variables obtained before treatment. Methods: A total of 43 patients with ST were randomized to the immediate-start group (n = 20) and delayed-start group (n = 23). All patients received physical therapy for 1 week (seven sessions). Each session lasted 60 min. The Visual Analogue Scale (VAS), Tinnitus Handicap Inventory (THI), and numerical pain rating scale (NPRS) scores were documented at baseline and after treatment (week 1) for all patients. For subjects in the immediate-start group, the THI, VAS, and NPRS scores were measured after therapy (weeks 6, 9, and 12, respectively). Medical history characteristic functional activity scale (HCFA) scores were measured at baseline to assess the association between somatic symptoms and tinnitus. Results: At week 1, VAS, THI, and NPRS scores of patients in the immediate-start group were improved by 1.25 ± 1.59, 11.10 ± 15.10, and 0.95 ± 1.54 points, respectively, and were significantly higher than those in the delayed-start group (p < 0.05). The change in VAS, THI, and NPRS scores in the treatment group was significantly positively correlated with the scores of the HCFA before treatment (r = 0.786, p < 0.001; r = 0.680, p = 0.001; r = 0.796, p < 0.001). There was no significant difference in THI, VAS, and NPRS scores among patients in the immediate-start group between weeks 1, 6, 9, and 12 after treatment (p > 0.05). Conclusions: Although more participants were necessary in the further study, the study implies that physical therapy can reduce physical pain, improve tinnitus symptoms, and quality of life in ST patients without hearing loss, and the short-term curative effect is stable, especially for tinnitus patients with clear somatic symptoms.