This paper reviews the Vermont Agency of Transportation Bridge Deck Membrane Evaluation Program begun in 1971 and discusses the field performance of 33 membrane systems over exposure periods of up to 11 years. Applications of deicing chemicals (sodium chloride) during the evaluation period have averaged 29.5 tons per two lane mile per year with accumulations totaling up to 123 pounds of chloride (Cl-) per linear foot of structure. Performance results are based upon the presence or absence of Cl-above base levels as determined by chemical analysis of over 1600 recovered concrete samples. The results show that almost without exception, the experimental systems have outperformed the Agency's original standard treatment of tar emulsion. When grouped by general type, the best performance has been provided by the standard preformed sheet membranes and thermoplastic systems. Although somewhat less successful, good performance has been provided by the polyurethanes, the NCHRP Project 12-11 recommended systems, and miscellaneous preformed systems. In general, the epoxy and tar emulsion systems were not considered successful although they have allowed only an average of 0.35 pounds of Clper cubic yard of concrete (#/cy) above base levels in the top inch of concrete as compared to an average of 6.97# Icy on exposed bridge decks over a similar evaluation period. Chloride contamination was detected in one sample or less on 33 percent of the 63 bridge decks under evaluation. Projections based upon performance results to date suggest a significant number of the membrane systems will provide protection from serious Cl-contamination for 50 years or longer.