Aim: To compare gingival marginal microleakage at the tooth structure / composite resin interface (closed sandwich technique) and tooth structure/glass ionomer interface (open sandwich technique).
Method: forty-five exfoliated sound human primary molars extracted at the time of shedding were selected. Class II cavity was prepared in each tooth. The teeth were randomly divided into three equal groups according to the restoration procedure, each group contained fifteen teeth; Group (I) Open Sandwich Technique, Group (II) Closed Sandwich Technique, and Group (III) Open Sandwich Technique without etching. All specimens were stored in artificial saliva for one week before thermocycling Microleakage at the gingival margin between the tooth structure and both glass ionomer liner and composite resin were assessed for the extent of dye penetration.
Results: In group (I); 66.7% of the samples had a score of 0 (no microleakage) 20% had a score of 1 and 20% had a score 3 and no sample had a score of 2. In group (II); 86.7% 
of the samples had a score of 0 (no microleakage) 20% had a score of 1 and 13.3% of the samples had a score 2. No samples had a score of 3 or 4. In group (III); No leakage (score 0) was detected in 60% of the samples while score "1" and score "2" microleakage was observed in 20%. Score "3" and score "4" dye penetration were not evident in any samples of group (III).
Conclusion: there was no significant difference between different groups. However, microleakage scores were lower in the open than in the closed sandwich technique. In addition, acid-etching on the glass ionomer liner did not eliminate microleakage.