Glass-ceramic materials, which are applied to dental and bone implantation, are increasingly important in surgery due to the improvement on controlled crystallization and the development of glass-ceramic materials. Apatite-containing glass-ceramics are potentially important for surgical implantation due to the close crystallographic and chemical similarity of apatite to bone tissue [1]. The crystalline phases occurring in these glass-ceramics include essentially, first, apatite, the major crystalline phase which provides the biocompatibility of the glass-ceramics, and, secondly, secondary crystalline phase(s), crystallized at higher temperature and providing the interesting mechanical properties such as machinability [2], high strength [3-5], etc. The apatite phase exhibits a prominent surface and a little lower bulk nucleation. The secondary crystalline phases often show only surface crystallization [3-6] with a surface layer grown perpendicularly to the free surface of the bulk glass, so large cracks are seen due to a large amount of directional volume change [5]. This limits the application of these glass-ceramics, and hence the method of glass powder sintering was used to eliminate the surface effects [3-5]. Nevertheless, the casting method for the fabrication of a bulk glass article is simpler than the glass powder sintering method. One way in which it is possible to resolve the problem of crack formation in the casting method is the addition of nucleating agent into glass-ceramics to elevate the internal nucleation of apatite, so that the growth of the surface layer would be retarded in its initial stage. In the work described in this letter, TiO2 was used as a nucleating agent for an MgO-CaO-SiO2-P205 glass, which possesses a composition locating at the eutectic point of the 3CaO.PzOs-CaO-MgO.2SiO2-SiO 2 system. The nominal compositions of the glasses are given in Table I. Well-mixed dried powders containing appropriate amounts of reagent-grade chemicals were melted in a platinum crucible for 2 h at 1450 °C and quenched by pouring on to a copper plate. The glasses were annealed at 685 °C for 2 h, then furnace-cooled to room temperature. A two-stage heat-treatment method was used to study the effect of heat-treatment temperature on