An alternative design to conventional class II cavity preparation for proximal carious lesions is the tunnel preparation. It preserves the marginal ridge intact, thus making it possible to maintain the natural contact relationship with the adjacent tooth and minimize tooth reduction. This in vitro study was purposed to evaluate the effect of the materials' elastic constants and shear-bond strength on the marginal ridge fracture resistance of teeth restored by the tunnel technique, and to find the materials of choice for tunnel restorations. $Resinomer^{(R)}$, $Ketac-silver^{(R)}$, $Miracle-Mix^{(R)}$, and Tytin were used as restorative material. The elastic constants of each restorative material were evaluated by ultrasonic pulse measurement. Young's modulus and bulk modulus of the restorative materials were evaluated in three specimens for each material type. The shear-bond strength of the restorative materials to the dentin surface was measured after thermocycling 400 times between 6 and $60^{\circ}C$, using ten specimens for each material type. For measuring marginal ridge strength, 60 sound extracted molar teeth were distributed into six groups by size. Sound molar teeth were used as a Control group and unfilled prepared teeth were grouped as Unrestored. Another four groups were named Resinomer group, Ketac-Silver group, Miracle Mix group, and Tytin group by type of restorative material. Tunnel cavity preparation was done with ' 1/2, 2, and 4 round burs in sequence. Initial access to proximal surface was made through an occlusal access preparation started at least 2mm from the marginal ridge, and the proximal opening was formed about 2.5mm below the marginal ridge. After restoration and thermocycling, marginal ridge strength was measured using a universal testing machine. The results were as follows: 1. The Young's modulus of $Tytin^{(R)}$ was 63.95 GPa, followed by $Ketac-Silver^{(R)}$ 27.60 GPa, $Miracle-mix^{(R)}$ 18.48 GPa, and $Resinomer^{(R)}$ 10.74 GPa showing significant differences between the groups(P<0.05). The bulk modulus of the materials showed the same order as Young's modulus. The value of $Tytin^{(R)}$ showed 59.57 GPa indicating that it will deform less than other materials under the same stress. It was followed by $Ketac-Silver^{(R)}$ 23.57 GPa, Miracle $Mix^{(R)}$ 12.50 GPa, and $Resinomer^{(R)}$ 11.60 GPa. 2. The Resinomer group had a shear-bond strength of 7.41 MPa which was significantly higher than those of the Ketac-Silver group (1.80 MPa) and the Miracle Mix group (2.84 MPa) (P<0.01). All the specimens of Tytin group detatched from the dentin surface during thermocycling. 3. The mean marginal ridge strength of the Unrestored group(46.14 kgf) was significantly lower than that of the Control group (84.24 kgf) (P<0.01). The marginal ridge strength of teeth restored by the tunnel technique was, in order, Ketac-Silver group 74.06 kgf, Miracle Mix group 73.36 kgf, Resinomer group 63.47 kgf, and Tytin group 58.76 kgf. The Ketac-Silver, Miracle Mix, and Resinomer groups showed no significant difference with the Control group (P>0.05), but the Tytin group showed significantly lower strength compared to the Control group(P<0.05). The results showed that the marginal ridge strength of the teeth restored by the tunnel technique was not significantly lower than that of sound teeth. They also demonstrated that the bonding strength of the restorative material to the tooth surface should be high and the modulus of elasticity should not be lower than that of the tooth in order to restore the marginal ridge strength to its natural condition.