AIM: Due to the immune privilege of the anterior eye chamber, the success rate of corneal transplantation can reach 90%. The aim of this study was to determine cytokine profile in aqueous humor of patients undergoing corneal transplantation, and to establish whether cytokine profile at the time of surgery influenced corneal graft outcome. METHODS: Proinflammatory (TNF-beta and IFN-gamma) and immunosuppressive (TGF-beta2) cytokine levels were measured in aqueous humor and serum of 44 patients. Non-inflammatory corneal diseases included keratoconus (n=8), bullous keratopathy (n=7), and stromal dystrophy (n=3). Inflammatory diseases included corneal scars (n=10), graft rejection (n=5), pending perforation (n=4), chemical burns (n=4), rejection/ uveitis (n=1), infectious keratitis (n=1), and perforated ulcer (n=1). Control aqueous humor and sera were obtained from cadavers without corneal pathology. RESULTS: The concentration of TGFbeta2 in the aqueous humor in non-inflammatory corneal diseases was similar to that of controls (2,605+/-204 pg/mL vs 2,200+/-100 pg/mL). In inflammatory corneal diseases, the concentration of TGFbeta2 in aqueous humor was significantly lower (1,400+/-375 pg/mL, p<0.001). TNF-beta was detected in the aqueous humor of 16 out of 26 patients with inflammatory corneal diseases and in all patients with stromal dystrophies, but was undetectable in cases of keratoconus and bullous keratopathy. Low levels of IFN-gamma were present in all aqueous humor samples. Patients' sera contained significantly less cytokine (up to 252 pg/mL) then their aqueous humor (p<0.001). We have set an arbitrary cut-off point for TGF-beta2 level in aqueous humor at 1,500 pg/mL and divided all investigated samples (from 44 patients and 10 controls) into two groups, one with high and the other with low TGF-beta2 concentration. The coefficient of contingency showed that patients with high TGF-beta2 concentration in their aqueous humor had significantly greater chance for graft acceptance than those with low TGF-beta2 concentration (p<0.001). CONCLUSION: High TGFbeta2 concentrations in the eyes without intraocular inflammation suggest its immunosuppressive role in human eyes. High concentration of TGFbeta2 (>1,500 pg/mL) was associated with graft acceptance. Also, absence of proinflamatory TNFbeta increased the graft acceptance, but independently from TGFbeta2.