Cellular processes like alternative splicing (AS) and proteolytic processing generate various protein isoforms from the same protein-coding gene. Studies using high-throughput sequencing indicate that about 90% of multi-exon genes in humans undergo AS. These AS events often occur in a tissue and developmental stage specific manner. Proteolytic processing is important for regulation of protein function and involved in processes like cell cycle regulation, apoptosis and protein degradation. Standard bottom-up proteomics involves digesting proteins into peptides. Since many peptide sequences match to several proteoforms, the information to which proteoform a given peptide belongs is lost at this step. Furthermore, there is an evolutionarily conserved preferential usage of lysine and arginine at splicing junctions. In combination with tryptic digestion, this impedes detection of splice junction-spanning peptides. Hence, the contribution of AS to protein diversity remains understudied. In this study we combined full-length mRNA sequencing (Iso-Seq) with proteomic analyses to obtain an integrated landscape of mRNA and protein isoforms in human RPE-1 cells. To overcome the limitations of bottom proteomics to study protein isoforms we resolved proteins by extensive protein-level fractionation by SDS-PAGE using the GELFREE 8100 fractionation system. After digestion of individual fractions and TMT labeling, the samples were combined and analyzed using conventional LC-MS/MS. A newly developed computational framework enabled us to automatically differentiate proteoforms from bottom up proteomics at a higher scale. As a result we established a reference set of ~45,000 full-length transcripts, ~32,000 ORFs and ~16,000 protein isoforms. Our data reveals multiple protein isoforms for many genes and provides an integrated landscape of mRNA and protein isoforms to reveal how transcriptional, translational and post-translational processes contribute to proteome complexity.