STATEMENT OF PROBLEM Most microleakage studies have used low molecular weight dyes or isotopes rather than clinically relevant materials, such as lipopolysaccharides or cell wall materials, that have been shown to provoke inflammatory reactions in the dental pulp. PURPOSE This study evaluated the leakage (diffusion) of lipopolysaccharide and dextran beneath cast gold crowns luted with 3 cements. MATERIALS AND METHODS Thirty extracted molars were prepared for crowns. Ten crowns with access ports (facial or lingual) were cast in gold and luted with zinc phosphate, glass ionomer, and an adhesive resin cement onto their preparations. Teeth and crowns with filters inserted into the ports were immersed in a solution of labeled macromolecules (TRITC-LPS, FITC-dextran) and evaluated for leakage at 2 weeks, and 1, 2, 3, 4, 5, and 6 months. Filters were retrieved and analyzed with fluorescent microscopy. RESULTS All filters retrieved from crowns luted with zinc phosphate, glass ionomer, and adhesive resin cements demonstrated no detectable leakage and were negative for both FITC-dextran and TRITC-lipopolysaccharide at all evaluation periods. CONCLUSION Leakage of lipopolysaccharides and dextran did not occur during the period of this study. For the length of this investigation, zinc phosphate, Ketac-Cem, and C&B-Metabond were equally effective at preventing leakage of detectable molecular concentrations of lipopolysaccharide and dextran to the level of the access ports.