The study investigated the erosion/abrasion-preventing potential of experimental NaF (1,400 ppm F-) and amine fluoride (AmF)/NaF/SnCl2/chitosan (1,400 ppm F-, 3,500 ppm Sn2+, 0.5% chitosan) toothpastes relative to placebo and SnF2 gel (970 ppm F-, 3,000 ppm Sn2+), and the impact of the demineralised dentine matrix on toothpaste effects. The study was a cyclic erosion/intervention experiment (10 days). Samples were stored in mineral salt solution either without or with collagenase (from Clostridium histolyticum type VII; 100 U/ml) for continuous removal of the organic matrix. To produce a comparable order of tissue loss, erosion was performed 6 × 30 s/day with 0.5% citric acid in the latter and 6 × 90 s/day with 1% citric acid in the former. Intervention was toothpaste slurry immersion (2 × 2 min/day); half of the samples were additionally brushed for 15 s within this time (brushing machine, load 200 g). Tissue loss was determined profilometrically (mean ± SD; µm). Tissue loss values (without/with brushing) for placebo, NaF, AmF/NaF/SnCl2/chitosan and SnF2 gel, respectively, were 11.6 ± 3.1/12.2 ± 2.5, 12.7 ± 3.1/10.7 ± 4.5, 8.7 ± 2.1/9.7 ± 2.1 and 8.8 ± 1.8/10.9 ± 1.8 in the presence of the organic matrix and 10.7 ± 3.2/11.9 ± 2.1, 8.2 ± 4.0/10.1 ± 4.1, 8.7 ± 2.9/9.1 ± 1.8 and 8.4 ± 1.9/7.5 ± 1.5 in the absence of the organic matrix. Relative to placebo, the NaF formulation had no significant effects; the AmF/NaF/SnCl2/chitosan formulation significantly reduced tissue loss between 20 and 25% except when applied without brushing in the absence of the organic matrix. The effects of the formulations were similar both in the presence and absence of the organic matrix. Sn2+/F--containing formulations have the potential to reduce erosion/abrasion even in the absence of demineralised collagen; seeking for more effective formulations is desirable.