A satisfactory drug release profile for gonadotropin-releasing hormone (GnRH) agonist drugs is high initial release followed by small amount of drug release per day. In the present study, three water-soluble additives (NaCl, CaCl2 and glucose) were selected to improve the drug release profile of a model GnRH agonist drug-triptorelin from PLGA microspheres. The pore manufacturing efficiency of the three additives was similar. The effects of three additives on drug release were evaluated. Under the optimal initial porosity, the initial release amount of microspheres containing different additives was comparable, this ensured a good inhibitory effect on testosterone secretion in the early stage. For NaCl or CaCl2 containing microspheres, the drug remaining in the microsphere depleted rapidly after the initial release. The testosterone concentration gradually returned to an uncontrolled level. However, for glucose containing microspheres, it was found that the addition of glucose could not only increase the initial release of the drug but also assist in the subsequent controlled drug release. A good and long-time inhibitory effect on testosterone secretion was observed in this formulation. The underlying cause why the incorporation of glucose delayed the subsequent drug release was investigated. SEM results showed that considerable pores in glucose containing microspheres were healed during the microspheres incubation. After thermal analysis, an obvious glass transition temperature (Tg) depression was observed in this formulation. As Tg decreased, polymer chains are able to rearrange at lower temperatures. This, morphologic change was reflected in the gradual closure of the pores, and is the likely reason that drug release slowed down after the initial release. HIGHLIGHTS The addition of glucose could not only increase the burst release of the drug but also delay the subsequent drug release. High initial burst and a sustained drug release helped obtain a good inhibitory effect on testosterone secretion. As Tg decreased, polymer chain was prone to rearrange. Morphologic change was reflected in the gradual closure of the pores. This was the reason that drug release slowed down after the initial burst.