Background. Enzootic bovine leukosis (EBL) remains a significant problem in cattle production. Agar gel immunodiffusion (AGID) is widely used for screening; however, diagnostic performance depends on test-kit manufacturing parameters. Aim. To develop and optimize manufacturing conditions for an AGID test system for serological diagnosis of EBL, ensuring reproducibility and compliance with World Organisation for Animal Health (WOAH) recommendations. Methods. BLV antigen was produced on a persistently infected FLK cell line and standardized to gp51 using WOAH reference serum E05. We optimized agarose concentration (0.8–2.0% in 0.2 M Tris, pH 7.2, with 8.5% NaCl), antigen dilutions (1:2–1:64), and control sera dilutions. Plates were incubated in a humid chamber at 20–27 °C with readings at 24–72 h. Validation was performed on bovine sera (positive/weak-positive/negative) and compared with commercial kits. Results. An agarose concentration of 1.0–1.2% provided clear, stable precipitin lines within 24–72 h, balancing diffusion rate and line quality. Optimal antigen dilutions were 1:4 (eliminating nonspecific lines with negative serum while maintaining high sensitivity) and 1:8. Positive control serum was optimal at 1:4–1:8; weak-positive control at 1:32 yielded a reproducible faint line suitable for sensitivity assessment. Negative serum produced no lines. Stable performance was achieved under specified storage and operating conditions; diagnostic characteristics were comparable to those of commercial tests. Conclusions. The optimized AGID test system demonstrates high reproducibility, specificity, and adequate sensitivity, aligns with international requirements, and is suitable for broad implementation in veterinary laboratories in Kazakhstan. Accounting for regional features further enhances its practical value.