Accurate isoform and alternative splicing detection are crucial for understanding transcriptional regulation and gene function. Long-read RNA sequencing (RNA-seq) technologies offer significant advantages over short-read RNA sequencing in capturing the full complexity of transcriptomes. Over the past decade, benchmarking studies have been crucial in evaluating isoform and AS detection workflows based on RNA-seq data. Different benchmarks have been developed to evaluate analytical factors with unique characteristics. However, the lack of comprehensive, biologically grounded, and reliable transcriptome-wide reference datasets presents a challenge for assessing and comparing the performance of various RNA-seq platforms and data analysis pipelines. As part of the Quartet Project, we established a set of publicly available RNA reference materials from four peripheral blood B lymphoid immortalized cell lines in a quartet family form the Taizhou Cohort of Fudan, including a pair of identical twin daughters. In this study, we developed the unique ratio-based reference datasets based on the Quartet RNA reference materials. The reference datasets were designed explicitly for benchmarking isoform and alternative splicing detection combining long- and short-read sequencing technologies to address previous challenges. The Quartet RNA reference datasets integrated multiple biological replicates, different sequencing technologies, platforms (PacBio, ONT, MGI), library preparation methods (cDNA and direct RNA), and various detection pipelines, capturing transcriptomic diversity and facilitating rigorous, transcriptome-wide evaluations. Extensive orthogonal validations confirmed the reliability of detected isoforms and splicing events in the reference datasets. The quality control framework and detection workflow based on the reference datasets can also facilitate developing and optimizing long-read sequencing technologies, ultimately advancing our understanding of transcriptomics. The Quartet RNA reference datasets will serve as a critical benchmark for the transcriptomics community, especially for isoform and alternative splicing detection using RNA-seq. By enhancing the accuracy and reliability of transcriptome analysis, our work will support various applications in biomedical research and precision medicine.