The aim of this experimental study was to evaluate the use of resorbable implants for the repair of nonloaded skeletal defects. Porous ceramic implants of alpha-TCP, of glass-ceramic, and of solid composite implants of glass-ceramic/polylactic acid 8 mm in diameter and 2 mm in thickness were fabricated and implanted pressfit into biparietal, full-thickness defects of the calvaria of 60 adult rats. Twenty rats received unfilled defects and served as controls. Fluorochrome labeling of bone formation was performed during the observation period. Five animals from each group were evaluated after 6, 13, 26, and 52 weeks. The control defects showed incomplete regeneration, with bone formation extending 1.66 mm, on average, into the defect after 52 weeks. In the group of alpha-TCP implants, histologic evaluation indicated that the bone formed during initial stages had undergone resorption later on, so that bone repair after 52 weeks was not significantly enhanced, with an average depth of 1.83 mm of bone ingrowth. The glass-ceramic implants exhibited extensive bone formation and nearly complete repair of the calvarial defect, with 3.90 mm of bone ingrowth into the implant pores. Degradation of the ceramic was nearly complete, with a few remaining particles surrounded by soft tissue. The composite implants showed a negligible bone ingrowth of 0.63 mm, on average. Soft tissue had invaded the polylactic acid implant body, but no bone formation had taken place at the surface of the embedded ceramic particles. Degradation of the polymer was not complete after 52 weeks. It is concluded that the balance between degradation and bone formation is delicate and that chemical events and cellular reaction during degradation may counteract complementary bone ingrowth.