Objective:To investigate the relationship between smoking status at presentation and the use of strabismus surgery in the management of patients with thyroid eye disease. Design:Retrospective review of a noncomparative series of patients with thyroid eye disease. Participants:All patients with thyroid eye disease under the care of a single consultant at Moorfields Eye Hospital between 1997 and 2002 (inclusive). Methods:Retrospective review of clinical case notes. Main Outcome Measures:Survival analysis of patients in cohort and the frequency of strabismus surgery in relation to smoking status at ophthalmic presentation. A subanalysis of patients who underwent orbital decompression and those that did not was undertaken. Results:Of 501 patients seen during the study period, 425 (85%) of 501 sets of notes were available for review, and initial smoking status was recorded for 89% (378/425) of patients, of whom approximately one half (196/378; 52%) were active smokers. Of the smokers, 51 (26%) of 196 underwent strabismus surgery, compared with only 19 (14%) of 138 nonsmokers at presentation. When adjusted for age, the hazards ratio of having strabismus surgery during management for smokers at presentation versus nonactive smokers was 2.19. In the group who did not undergo orbital decompression, this hazard ratio increased to 4.86. Conclusions:Within this thyroid eye disease cohort, the proportion of smokers at presentation was much larger than that of the general population. There was an increased use of strabismus surgery in active smokers at presentation than in nonactive smokers. This finding was independent of the orbital decompression surgery. The results are consistent with those of previous reports of more severe thyroid eye disease in smokers and raise the possibility that smoking cessation early in the disease may reduce the severity of the changes and the number of rehabilitative strabismus operations needed.