Purpose: To establish a rabbit model for keratoconjunctivitis sicca (KCS) to study autologous submandibular gland transfer for treating severe KCS. Methods: In 2 groups of 10 rabbits, left eyes were operated and right eyes were controls. In the trichloroacetic acid-treated group, the lacrimal and harderian glands and nictitating membrane were removed surgically; the palpebral and bulbar conjunctiva were swabbed with 50% trichloroacetic acid. In the non-trichloroacetic acid-treated group, the lacrimal and harderian glands and nictitating membrane were surgically removed. The Schirmer test was performed preoperatively and 1, 2, 3, and 4 months postoperatively. Corneal densities of rose bengal and fluorescein staining were scored every month postoperatively. At 4 months, the cornea and bulbar conjunctiva were removed from operated and control eyes for histopathology. The upper bulbar conjunctiva was used to determine goblet cell density. Results: Compared with preoperative conditions, tear secretion of operated eyes significantly decreased in both groups postoperatively, then gradually increased. Scores for corneal rose bengal and fluorescein staining were higher and conjunctival goblet cell density was lower in the operated eyes than in control right eyes in both groups, but no significant difference was found between the operated eyes of the two groups. Inflammatory histopathologic changes of the cornea and conjunctivawere not found in either of the eyes in the two groups. Conclusions: A new rabbit model for KCS could be created by either of these methods. Experimental KCS with reduction of tear production was possible with surgical ablation of the lacrimal and harderian glands and nictitating membrane. It is unnecessary to apply trichloroacetic acid to burn the conjunctiva. Our modified incision better exposed the surgical field.