A very small portion of the genome codes for genes. However, not only protein-coding DNA is relevant to understand the functioning of a cell. Non-coding RNA transcripts (ncRNAs) have been shown to play a central role in post-transcriptional regulation. Some of these mechanisms rely on the binding of ncRNAs to RNA-binding proteins (RBPs). A subset of ncRNAs, called long non-coding RNAs (lncRNAs), have been poorly characterized to date, due to experimental technical limitations that have recently been overcome. We present a methodology that aims to discover high affinity RBP binding sites in lncRNA transcripts using known RBP motifs as an input and exploiting computational predictions of the RNA secondary structure. The proposed methodology has been applied to well-studied lncRNAs with a known relation with cancer, such as HOTAIR and ANRIL. Results are shown to be consistent when validated against a lncRNA-protein interaction database.