One hundred Babcock 308 female-layer chickens were randomly divided into five groups of 20 each. Groups 1, 2, 3 and 4 were different positive control groups of immunized chickens with commercial trivalent mineral oil vaccine, and prepared bacterins of Avibacterium paragallinarum serovars A (221), B (0222) and C (Modesto), respectively. The chickens in Group 5 were assigned as a negative control and immunized with PBS. The serum from Groups 1–5 at 4 weeks after the first vaccination were used to calculate the sensitivity and specificity of the newly developed indirect enzyme-linked immunosorbent assay (I-ELISA) method. Forty negative control sera (taken before vaccination) were used to evaluate the cut-off value of the I-ELISA against each serovar of A. paragallinarum under optimal conditions. The cut-off values of serovars A, B and C, calculated by the mean optical density of all the negative sera plus three standard deviations were 0.334, 0.484 and 0.678, respectively. The efficacy of the developed I-ELISA showed 100% sensitivity for all three serovars of coating antigen but with a low specificity of 30% for all three serovars because of the high cross reactivity among serovars. Nevertheless, the serovar A I-ELISA gave a higher response to serovar A antibodies than to the other two heterologous serovars (P < 0.05). In contrast, the I-ELISA results for B and C did not show any significant difference between the homologous and heterologous serovars. In conclusion, this newly developed I-ELISA could be an alternative method for differentiating between A. paragallinarum-free chickens and those that have received either a vaccination and/or a challenge exposure.