Ceramometal crowns are common restorations in fixed prosthodontics because of their casting accuracy, the high strength properties of the metal, and the cosmetic appearance of porcelain. However, deterioration of the initial fit of the metal coping has been observed after the porcelain firing cycle. The distortion due to repeated firing makes it difficult to fit crown margin and elicits microleakage. The major causes of distortion are the residual stress that accumulate during wax-up, casting, cold work and the induced stress caused by the mismatch of porcelain-metal thermal contraction. This study examined the marginal fit changes of metal copings in relation to repeated firing and the effects of heat treatment that reduce the distortion resulted from residual stress. The marginal changes of the copings that were treated with conventional method and those treated with heat before repeated firing, were evaluated. The metal die which represented preparations of a maxillary central incisor was fabricated, and 45 wax patterns were cast with nonprecious metal alloys. The heat treatment of each group was performed as follows. Group 1(control) : Casting - Devesting - Cold work - Firing Group 2 : Casting - Heat treatment - Devesting - Cold work - Firing Group 3 : Casting - Devesting - Cold work - Reinvesting - Heat treatment - Devesting - Firing The copings were fired 3 times. After each firing, the marginal fit changes were measured with inverted metallurgical microscope at the 4 reference points located at labial, lingual, and both proximal surface. Measurements were compared, and statistically analyzed. The results were as follows ; 1. In all groups, the highest value of marginal fit changes of the copings studied were found after the first firing cycle. 2. When the distortion of each experimental group at the first firing cycle were compared, group 1 exhibited the greatest changes($20-27{mu}m$), followed by group 2($9-13{mu}m$), and group 3($8-10{mu}m$). 3. The copings treated with heat before devesting(group 2) revealed significantly smaller marginal fit changes than the copings treated with conventional method(group 1). (p<0.01) 4. The copings treated with heat after reinvesting(group 3) revealed significantly smaller marginal fit changes than the copings treated with conventional method(group 1). (p<0.01) 5. No siginificant differences in marginal fit changes were found between the copings treated with heat before devesting(group 2) and the copings treated with heat after reinvesting(group 3). (p>0.01)