A variety of laboratory tests have been developed to assist in predicting the clinical performance of dental restorative materials. Additionally, more than one methodology is in use for many types of tests performed in vitro. This project assessed and compared results derived from two specific laboratory testing methods, one for bond strength and one for microleakage. Seven multi-purpose dental adhesives were tested with the two methodologies in both amalgam and resin composite restorations. Bond strength was determined with a punch-out method in sections of human molar dentin. Microleakage was analyzed with a digital imaging system (Image-Pro Plus, Version 1.3) to determine the extent of dye penetration in Class V preparations centered at the CEJ on both the buccal and lingual surfaces of human molar teeth. There were 32 treatment groups (n = 10); seven experimental (dental adhesives) and one control (copal varnish, 37% phosphoric acid) followed by restoration with either amalgam or resin composite. Specimens were thermocycled 500 times in 5 degrees and 55 degrees C water with a one-minute dwell time. Bond strength and microleakage values were determined for each group. ANOVA and Student-Newman-Keuls tests demonstrated an interaction between restorative material and adhesive system with a significant difference among adhesives (p < 0.05). Using a multi-purpose adhesive system resulted in both a statistically significant increase in bond strength and a statistically significant decrease in extent of microleakage (p < 0.05). The effect of the adhesive upon both microleakage and bond strength was greater in the resin composite restorations than in the amalgam restorations. Bond strength testing was more discriminating than microleakage evaluation in identifying differences among materials.