Caprine tuberculosis (TB) is a zoonotic disease caused by members of the Mycobacterium tuberculosis complex. TB eradication programs in goats are based on the single and comparative intradermal tuberculin tests (SITT and CITT, respectively). Antibody-based diagnostic techniques have emerged as potential diagnostic tools for TB. P22 ELISA has been previously evaluated using samples collected after the intradermal tuberculin tests to maximize the sensitivity, a phenomenon known as booster effect. However, there is no information available on whether the use of this diagnostic strategy could lead to a decrease of its specificity (Sp). The aim of the present study was to elucidate the interference effect of a recent CITT on the Sp of the P22 ELISA in serum and milk samples collected at different times after the CITT from a TB-free herd (n = 113). The number of reactors to P22 ELISA was significantly higher (p < 0.01) on serum samples collected 15 days post-CITT compared to day 0, showing a decrease in Sp from 99.1% (95% CI; 95.2-99.8%) to 88.5% (95% CI; 81.3-93.2%). The number of reactors and the quantitative values of P22 ELISA were significantly higher (p < 0.01) in serum samples compared to milk. No significant (p > 0.05) changes in the Sp of the P22 ELISA were observed throughout the different time samplings using milk No significant (p > 0.05) changes were observed on days 30 and 60 post-CITT. In conclusion, the booster effect strategy may significantly decrease the Sp of P22 ELISA in TB-free herds when serum samples are used but not when milk is tested.