Prediction of the effects of splice-site variations by sequence analysis is difficult. In this study we provide the means for a rapid evaluation of the potential for splice-site mutations to interfere with RNA processing. The system may be useful in reverse genetics or mapping studies when isolation and characterization of mRNA is arduous or not possible. In the assay we cloned wild-type and mutant sequences of murine splice-site mutations into an exon-trapping vector and characterized splicing of both recombinant transcripts in a transient cell culture system. Results from this artificial assay were compared with in vivo data from the respective mouse models. We found that the exon-trapping system allows one to confidently predict whether a splice-site variation is going to have a splicing effect in vivo, but the system does not always reflect in vivo splicing in detail. In summary, the exon-trapping system is a reliable and easy-to-use tool for a first evaluation of splice effects.