To investigate the effect of doxycycline on the nucleolar organizing regions and a-smooth muscle actin expression in bovine corneal myofibroblasts in vitro and assess its contribution to ocular surface repair mechanisms. Cell culture and identification: bovine corneal fibroblasts were cultured after the stroma was incubated in 1.0 and 2.0 g/L type I collagenase in two stages.Isolated cells were plated at mantaryay culture flask in 10% of BSA RPMI-1640. Vimentin and alpha-smooth muscle actin (α-SMA) organization were evaluated by immunocytochemistry. The cells staining positive for Vimentin and α-SMA indicated the presence of corneal myofibroblasts. Bovine corneal myofibroblasts were treated with different concentrations of doxycycline (10, 20, 40, 60, 80 mg/L) , a bland control group and the dexamethasone group (120 mg/L) were set up, each group had 30 cases. The argyrophilic nucleolar organizing regions (AgNOR) staining and the immunohistochemistry for α-SMA were performed when the cells were treated for 24 hours and 48 hours. The AgNOR count (Ag-c), AgNOR area (Ag-a) and the expression of α-SMA in the bovine corneal myofibroblasts among each experiment group and control group were compared using one-way ANOVA, further pairwise comparisons using Independent-Samples t test. Cell culture techniques were successfully used to establish a method for the isolation and culture of bovine corneal myofibroblasts. Microscopic examination and immunohistochemical staining confirmed that the cells cultured were bovine corneal myofibroblasts. The Ag-c and Ag-a of bovine corneal myofibroblasts progressively decreased as the concentrations of doxycycline was increase. 24 h:bland control group Ag-c was 6.40 ± 0.6, 60 mg/L doxycycline group Ag-c was 2.23 ± 0.43;bland control group Ag-a was (34.80 ± 2.36) µm(2), 60 mg/L doxycycline hormone group Ag-a was (19.91 ± 2.15) µm(2). 48 h: bland control group Ag-c was 7.27 ± 0.6,60 mg/L doxycycline hormone group Ag-c was 2.80 ± 0.76, bland control group Ag-a was (36.27 ± 1.99) µm(2), 60 mg/L doxycycline group Ag-a was (13.75 ± 2.09) µm(2). The differences were statistically significant: in the same time intervention (FAg-c 24 h = 252.55, FAg-a 24 h = 202.16, P < 0.05, FAg-c 48 h = 169.38, FAg-a 48 h = 853.23, P < 0.05), in the same concentrations intervention (tAg-c = 6.98, tAg-a = 11.62, P < 0.05). And 60 mg/L of doxycycline had an obviously inhibitory action as 120 mg/L dexamethasone in the same treated hours (dexamethasone group Ag-a 24 h = 30.56 ± 3.66, dexamethasone group Ag-a 48 h = 28.35 ± 1.23 ),the differences were not statistically significant (tAg-a 24h = 1.182, P = 0.242,tAg-a 48 h = 0.21, P = 0.832). As the concentrations investigated, doxycycline can inhibit the expression of α-SMA in the bovine corneal myofibroblasts (189.90 ± 7.48, 140.20 ± 7.79, 113.20 ± 8.98, 98.00 ± 3.50, 85.50 ± 4.99), the difference was statistically significant (F = 761.79, P = 0.00). While dexamethasone had no significant role in the expression of α-SMA (bland control group was 225.10 ± 6.74, the dexamethasone group was 228.50 ± 7.12), and the statistically difference was not obvious (t = 1.096, P = 0.287). As the concentrations of doxycycline was increased from 10 mg/L to 80 mg/L, the AgNOR count and AgNOR area of bovine corneal myofibroblasts can be significantly reduced in vitro. Compared with dexamethasone, doxycycline significantly suppressed the expression of α-SMA in bovine corneal myofibroblasts in a dose-dependent positive trend.