Protein–protein interactions (PPIs) are crucial in many biological processes. The first step towards the molecular characterisation of PPIs implies the charting of their interfaces, that is, the surfaces mediating the interaction. To this end, we present here iFrag, a sequence-based computational method that infers possible interacting regions between two proteins by searching minimal common sequence fragments of the interacting protein pairs. By utilising the sequences of two interacting proteins (queries), iFrag derives a two-dimensional matrix computing a score for each pair of residues that relates to the presence of similar regions in interolog protein pairs. The scoring matrix is represented as a heat map reflecting the potential interface regions in both query proteins. Unlike existing approaches, iFrag does not require three-dimensional structural information or multiple sequence alignments and can even predict small interaction sites consisting only of few residues. Thus, predicted interfaces range from short fragments composed of few residues to domains of proteins, depending on available information on PPIs, as we demonstrate in several examples. Moreover, as a proof of concept, we include the experimental validation on the successful prediction of a peptide competing with the aggregation of β-amyloid in Alzheimer's disease. iFrag is freely accessible at http://sbi.imim.es/iFrag.