The aim of this study was to assess the protective effect of experimental solutions containing chitosan at different viscosities with or without fluoride (TiF 4 /NaF) on dentin loss in vitro. Bovine dentin samples (n = 15) were prepared and allocated to one of the following treatments: (i) 0.5% chitosan (500 mPas); (ii) 0.5% chitosan (2,000 mPas); (iii) 0.042% NaF and 0.049% TiF 4 ; (iv) as (iii) with addition of 0.5% chitosan (500 mPas); (v) as (iii) with addition of 0.5% chitosan (2,000 mPas); (vi) commercial solution with SnCl 2 /AmF/NaF (positive control); or (vii) deionized water (negative control). The samples were submitted to pH cycling for 7 d (0.1% citric acid, 4 9 90 s d À1). The treatment was applied once a day for 30 s. The dentin loss was quantified using a contact profilometer. Three samples per group were evaluated using scanning electron microscopy. The dentin loss (lm) was submitted to ANOVA and Tukey's test for differences between treatments. Among the treatments tested, only chitosan 500 mPas was able to statistically significantly reduce the dentin loss compared to the negative control, being similar to the positive control. TiF 4 /NaF, whether with or without chitosan, had no protective effect. Chitosan 500 mPas and SnCl 2 /AmF/NaF solutions have comparable protective effect against dentin erosion in vitro.