The classical adjuvants alum and Freund's Incomplete Adjuvant (IFA) are frequently used as references for the design of new adjuvants and antigen delivery systems, e.g., microspheres (MS). Poly(dl-lactic-co-glycolic acid) (PLGA) MS have been proposed for delivering antigen booster doses in vivo after a single injection. However, as antigen release kinetics from conventional adjuvants are generally unknown, it appears presumptuous to propose a desired antigen release pattern from PLGA MS. Therefore, we have studied the tetanus toxoid (Ttxd) in vitro release from alum, IFA formulations and MS in four different test systems. The results showed a stronger Ttxd association to alum than to IFA, and the release from both formulations lasted between 3-9 days. The total of ELISA-responsive antigen released was 60-85% of the actual dose. Both the total amount and the prolongation of release depended on the Ttxd dose. Furthermore, the incomplete in vitro release of Ttxd from the adjuvants and also from PLGA 50:50 MS was shown to be partly due to experimental conditions. Typically, Ttxd adsorbed on the glass vials used for the release test and also on the surface of the PLGA 50:50 MS, wherefrom it was released. In conclusion, the test system depending rate and quantity of release observed evidence the limitations of in vitro release data. Finally, for mimicking conventional vaccination schedules, i.e. injections typically at time points 0, 1, 3, and 12-24 months, PLGA MS should release antigen doses at the corresponding time points, and the release pulse should only last for a few days.