Porcine circovirus type 2 (PCV2) is associated with a number of diseases and syndromes, collectively referred to as porcine circovirus-associated disease. The main objective of this study was to define some in vitro correlates of protection after injection of inactivated PCV2 vaccines with a defined antigen mass. Twelve pigs were vaccinated with three different doses of inactivated, whole-virus antigen (211-844 ng), while four animals were injected with a commercial vaccine (positive control) and four other pigs were mock-vaccinated with phosphate-buffered saline (PBS) in the same oil emulsion. Four weeks later, they were intranasally challenged with 2 × 10(5) TCID50 of a PCV2a strain. Antibody was measured in blood and oral fluids by enzyme-linked immunosorbent assay (ELISA) and a neutralization assay. PCV2 was quantified in serum by real-time polymerase chain reaction for ORF2 gene. PCV2-specific cell-mediated responses were investigated by an IFN-γ release assay in whole blood, IFN-γ ELISPOT, and lymphocyte proliferation (Ki-67 and BrDU assays). All the vaccines under study but mock provided complete or incomplete protection from PCV2 infection in terms of post-challenge viremia. Serum antibody titers (ELISA and neutralizing) after vaccination were not correlated with protection, as opposed to the early neutralizing antibody levels of vaccinated pigs at day 7 after infection. Cell-mediated immune parameters showed a good correlation with vaccine efficacy. In particular, the IFN-γ release assay at 3 weeks after vaccination was an effective marker for predicting protection. All control pigs always tested negative in assays of cell-mediated immunity. Our results outline in vitro testing procedures toward reduced animal usage in the control of PCV2 vaccine batches.