Splicing alterations are being increasingly reported to cause human diseases. However, predicting beforehand whether a given mutation might lead to aberrant splicing is often hampered by insufficient knowledge of which cis-acting sequences affect exon recognition. To better define these sequences, experimental methods have been developed that integrate pre-mRNA splicing with sequence selection assays. Recently, a novel in vivo selection method based on a partially randomized full-length exon has enabled the identification of new splicing-controlling elements in SMN exon 7. Skipping of this exon results in an aberrant protein and is involved in proximal spinal muscular atrophy (SMA). Hopefully, this approach will provide novel targets for nascent RNA molecular medicine approaches for the recovery of abnormal pre-mRNA splicing events.