Air-cooled slag showed grindability approximately twice as good as that of water-cooled slag. While the studied water-cooled slag was composed of glass as constituent mineral, the air-cooled slag was mainly composed of melilite. It is assumed that the sulfur in air-cooled slag is mainly in the form of CaS, which is oxidized into $CaS_2O_3$ when in contact with air. $CaS_2O_3$, then, is released mainly as $S_2O{_3}^{2-}$ion when in contact with water. However, the sulfur in water-cooled slag functioned as a constituent of the glass structure, so the$S_2O{_3}^{2-}$ ion was not released even when in contact with water. When no chemical admixture was added, the blended cement of air-cooled slag showed higher fluidity and retention effect than those of the blended cement of the water-cooled slag. It seems that these discrepancies are caused by the initial hydration inhibition effect of cement by the $S_2O{_3}^{2-}$ ion of air-cooled slag. When a superplasticizer is added, the air-cooled slag used more superplasticizer than did the blast furnace slag for the same flow because the air-cooled slag had higher specific surface area due to the presence of micro-pores. Meanwhile, the blended cement of the air-cooled slag showed a greater fluidity retention effect than that of the blended cement of the water-cooled slag. This may be a combined effect of the increased use of superplasticizer and the presence of released $S_2O{_3}^{2-}$ ion; however, further, more detailed studies will need to be conducted.