To determine the effect of resin based sealer on retention of casting cemented with three different luting agents. 55 extracted molar teeth were prepared with a flat occlusal surface, 20° taper and 4 mm axial height. The axial surface of each specimen was determined. The specimen were then distributed into five groups based on decreasing surface area, so each cementation group contained 11 specimens with similar mean axial surface area. A two-step, single bottle universal adhesive system (One-Step-Resinomer, Bisco) was used to seal dentin after the tooth preparation. Sealer was not used on the control specimens except for the modified-resin cement (Resinomer, Bisco) specimens that required use of adhesive with cementation. Using ceramometal (Wirobond(®), BEGO), a casting was produced for each specimen and cemented with either zinc phosphate (Harvard), glass ionomer (Vivaglass) or modified resin cement (Resinomer) with single bottle adhesive. All the castings were cemented with a force of 20 kg. Castings were thermal cycled at 5 and 55 °C for 2,500 cycles and were then removed along the path of insertion using a universal testing machine at 0.5 mm/min. A single-factor ANOVA was used with a = 0.05. The nature of failure was also recorded. The mean stress removal for non sealed zinc phosphate, sealed zinc phosphate, non sealed glass ionomer, sealed glass ionomer and modified resin cement was found to be 3.56, 1.92, 2.40, 4.26, 6.95 MPa respectively. Zinc phosphate cement remained principally on the castings when the tooth surface was treated with the sealer and was found on both the tooth and the casting when the sealer was not used. Fracture of root before dislodgement was seen in 9 of 11 specimens with modified resin cement. Resin sealer decreases the retention of the castings when used with zinc phosphate and increases it when used with glass ionomer cement. The highest mean dislodgement force was measured with modified resin cement.