OBJECTIVE A significant factor in achieving satisfactory adhesion of restorative resins to dentine substrate is the method by which the dentine surface is treated before an adhesive is applied. The aim of this study was to evaluate the effect of deproteinization on microleakage around gingival third resin restorations. METHODS Standardised Class V preparations were made on randomly selected intact upper and lower human molars. These were treated in one of five ways (no treatment, enamel etch only, total etch, total etch followed by deproteinization, and deproteinization only) and then adhesively bonded using either an acetone or ethanol based bonding system. The samples were first immersed in 2% methylene blue dye and then 35% nitric acid, for 72 hours each. The solutions were filtered and centrifuged, and the supernatant was used to determine absorbance in a spectrophotometer at 670 nm. The results were recorded as a measure of transmission of light of the test solutions. RESULTS The results were subjected to multiple comparisons amongst groups, using anova. There was a statistically significant difference between all treatment groups for the two different bonding systems used. The experimental groups, total etch alone and total etch followed by deproteinization showed statistically significant differences, as compared to all other groups. However, although the total etch group showed a decrease in microleakage, when compared to the total etch followed by deproteinization group, this was not statistically significant. CONCLUSIONS Within the limitations of this study, collagen removal may be important to reduce microleakage whilst using acetone based adhesive systems and it may not influence the amount of microleakage for ethanol or water based adhesive systems.