The purpose of this study was to evaluate the microleakage electrochemically using different retrograde filling materials, different root resection angle, and different cavity preparation instruments. 104 extracted single-rooted teeth were selected for this study. 100 teeth were used as experimental groups and four was used as controls. Anatomical crowns were resected, root canals were prepared, and the apical 2 mm of roots were removed. The experimental roots were randomly divided into five equal groups. Experimental groups : Group 1. no bevel, cavity preparation with ultrasonic instruments, amalgam filling Group 2. no bevel, cavity preparation with ultrasonic instruments, SuperEBA cement filling Group 3. no bevel, cavity preparation with ultrasonic instruments, desiccated ZOE filling Group 4. $45^{circ}$ bevel, cavity preparation with ultrasonic instruments, amalgam filling Group 5. no bevel, cavity preparation with conventional bur, amalgam filling Microleakage was measured once a day for 30 days using electrochemical method and were analyzed statistically. The results were as follows : 1. The group with Super EBA cement filling showed the least marginal leakage from second to fourth day(p<0.05), there was no significant difference between the group with amalgam filling during eighth to eighteenth day(p>0.05), but after the nineteenth day here was a higher marginal leakage than the group with amalgam filling(p<0.05). 2. The group with desiccated ZOE filling demostrated that the highest marginal leakage, started on the eighth day(p<0.05). 3. The group using ultrasonic instrument showed lower marginal leakage than the group using bur until the nineteenth day(p<0.05), but there was no significant differnce with the group using bur after twentythird day(p>0.05). 4. The group without bevel showed lower marginal leakage than the group with bevel (p<0.05). 5. Whether bevel or nor had much more effect on marginal leakage than with cavity perparation instrument when the cavity was retrogrdefilled with amalgam(stepwise regression).