The purpose of this investigation was to determine whether non-setting calcium hydroxide [Ca (OH)2] cement placed in the root canal system of premolar teeth would affect the subsequent microleakage of a glass-ionomer restoration (GIC). Following selection, 62 human premolar teeth extracted for orthodontic reasons were accessed and root canals prepared according to a standardized procedure. The specimens were then allocated randomly into two major groups each of 30 teeth. Two other teeth were used as a positive and a negative control. The control group was restored with glass-ionomer cement following drying of the canal and placement of a cotton wool pledget. The test group had all canals dressed with non-setting Ca(OH)2 and then was subdivided, one set (n = 22) being restored following conditioning of the access cavity margins, the other (n = 8) having the margins cleaned with a hand excavator. Samples were assessed for microleakage using a two-point scoring system (leakage or no leakage) in conjunction with a clearing technique using AgNO3. Using Fisher's exact test, a statistically significant difference was found between the control and test groups (P < 0.05) but there was no significant difference between the excavated and conditioned cavities (P=0.55). It is concluded that contamination of access cavity margins with Ca(OH)2 during medication of a root canal interferes with the bond of GIC, resulting in increased microleakage in vitro.