In this study we present a method for relative quantification of splicing events from RNA-sequencing data called Altrans. Our approach is an annotation-based method, which makes the least number of assumptions from the annotation. To this end we chose to simplify the problem and quantify relative frequencies of observed exon pairings in RNA-sequencing data for all categories of splicing events. This approach assumes only correct knowledge of the exons in the transcriptome and is agnostic to the isoform structures defined in an annotation, which would, in theory, make it more accurate and sensitive in the presence of unknown isoforms. We tested the performance of Altrans versus two well-established transcript quantification methods, Cufflinks11 and MISO,13 and benchmarked our method in two ways. First, we conducted a simulation study and assessed the concordance of the measured quantifications by each method with the simulated quantifications. Second, we assessed the relative power of discovering alternative splicing quantitative trait loci (asQTLs) for each method. For the asQTL analyses, we chose the Geuvadis dataset, since it was, at the time of analyses, the largest publically available population-based RNA-sequencing study. The Geuvadis dataset comprises 462 individuals in the 1000 Genomes project14 from five populations—the CEPH (CEU), Finns (FIN), British (GBR), Toscani (TSI), and Yoruba (YRI)—and contains data for whole-genome DNA sequencing and deep mRNA sequencing in the lymphoblastoid cell line (LCL)7 and is thus an ideal dataset for our purposes.