This study was conducted to determine the feasibility of injecting the somatostatin analogue, octreotide acetate (OA), into the vitreous cavity. Previous work suggests that octreotide effectively inhibits angiogenesis in vitro, thus its use in vivo may slow the progression of proliferative eye disease. Fifty micrograms of aqueous OA in 50 microliters aqueous solution was injected into the mid-vitreous of kitten eyes (n = 6), and OA levels were monitored over 4 days. A long-acting release form of octreotide (OA-LAR) was also injected into the mid-vitreous of rabbit eyes at doses of 0.36 (n = 16), 1.1 (n = 1), 2.1 (n = 1), 4.05 (n = 1), 8.2 (n = 1), and 36 mg (n = 3) in solution; and octreotide concentrations were measured at various time points over 42 days. OA concentrations were determined by a highly specific radioimmunoassay. Aqueous octreotide was eliminated rapidly (t1/2 = 16 hours) from the vitreous of the kitten eye, with only negligible amounts recoverable 4 days post-injection. In the long-acting form, OA in the rabbit eye reached peak levels at 28 days. By 42 days, OA levels had declined to the 14-day level. Doses of OA-LAR of 1.1 mg or less produced no gross evidence of clinical toxicity and elicited no grossly visible ocular side effects. Doses greater than 1.1 mg produced significant toxicity, including cataracts and rubeosis. The 28-day peak release for long-acting OA implies that monthly intravitreal injections could provide continual high levels of OA. Intravitreal injection of long-acting OA provides sustained, high concentrations of drug, and deserves further study as a potential treatment of proliferative eye diseases.