This study aimed at developing phosphoric acids modified with different polyphenol-rich plant extracts and to verify their effect on the 24h shear bond strength (SBS) to non-eroded and eroded dentine. In this view, the groups contained experimental phosphoric acid (PA), either plain or modified with polyphenol-rich plant extracts, and a commercially available phosphoric acid was used as positive control. The groups were: 1) LabPA (37% phosphoric acid prepared in the laboratory); 2) PA-GSE (LabPA+2% Grape seed extract); 3) PA-Blueberry (LabPA+2% Blueberry extract); 4) PA-Cranberry (LabPA+2% Cranberry extract); 5) PA-GT (LabPA+2% Green tea extract); 6) CommPA (commercial PA, Kerr Gel Etchant, 37.5% phosphoric acid). For SBS, a total of 192 dentine specimens were prepared from sound extracted permanent human molars, by grounding the teeth until the mid-coronal dentine. All specimens underwent a standard smear layer, half of them were straight restored (non-eroded dentine, n=96) and the other half underwent erosion challenge (eroded dentine, n=96; 10 cycles of 1h exposure to human saliva followed by 5 min immersion in 1% citric acid). Consequently, 96 non-eroded and 96 eroded specimens were etched (15s) with the PAs, according to the experimental groups (n=16 specimens/group) and subjected to an etch-and-rinse adhesive system, followed by standardized restoration with a composite resin. After 24 h incubation in humid chamber, the specimens were subjected to SBS using a universal testing machine. The failure mode of each specimen was determined under a stereomicroscope and described qualitatively. Data was analysed with Q-Q plots, Shapiro-Wilk and Levene’s tests for normality and equality of variances, and General Linear Model (GLM) with SBScalc as outcome variable, and “Acid Group” and “Presence of Erosion” as factors. Both factors individually caused an effect to the outcome (p