A slag sample was produced during laboratory reduction tests of a steel slag for metal recovery. The reduction of the metal oxides changed slag basicity (CaO/SiO2) to about 1.6 and led to the formation of g-phase of dicalcium silicate, g-Ca2SiO4, which resulted in disintegration of the slag sample. A modification study of the slag sample after reduction was performed. The study aimed at acquiring data and fundamental information for choosing environmentally friendly and cost effective methods for preventing slag disintegration in order to increase slag utilization. Effects of some additives, containing SiO2 or P2O5, on the formation of the g-Ca2SiO4 were examined via slag melting tests. The slag was also modified by fast cooling, via air granulation using a laboratory granulation system. In these modification tests, a laboratory induction furnace was used to melt the samples at approximately 1700°C. Samples obtained before and after the modification tests were characterized. The study results showed that g-Ca2SiO4 and fines did not appear by adding MCP-F or a product from iron ore processing to increase P2O5 content in the reduced slag from a very low level, 0.07%, to around 0.3% or higher. A sand addition of 5.12% in the reduced slag could decrease slag CaO/SiO2 value to 1.34, which prevented Ca2SiO4 formation and slag disintegration. The reduced slag was cooled quickly via air granulation, which prevented formations of g-Ca2SiO4 in slag granules. Based on results from the melting tests and other literature, fundamental aspects regarding slag modification and treatment are discussed, along with utilization of the slag after the oxide reduction and modification.