Pancreatic cancer is highly malignant with a dismal outcome. Current treatment protocols based on cohort studies have limited efficacy in improving its prognosis. This may be caused by the heterogeneity of pancreatic cancer among individuals, which leads to different biological properties and different therapeutic outcomes. The rapid development of precision medicine may bring hope for pancreatic cancer. In recent years, some progress has been made in in areas such as targeted therapy, molecular subtyping and personalized tumor model for pancreatic cancer, which enable the application of precision treatment strategies in pancreatic cancer. The four major driver mutations (KRAS/TP53/CDKN2A/SMAD4) play an important role in the development of pancreatic cancer, but there are no targeted drugs available in clinical practice. A phase II clinical trial demonstrated that AMG510 was effective in solid tumor patients with KRAS G12C mutation, but the KRAS G12C mutation frequency was relatively low in pancreatic cancer. The POLO trial suggested that olaparib can improve progression-free survival in metastatic pancreatic cancer patients with germline BRCA mutations. Based on this result, NCCN guideline recommended olaparib for maintenance treatment of advanced pancreatic cancer with germline BRCA1/2 mutations. In addition, the phase II KEYNOTE-158 study suggested that pancreatic cancer patients who were deficient in DNA mismatch repair (dMMR) and have high microsatellite instability may benefit from anti-programmed death-1 therapy with pembrolizumab. In general, the actionable targets for pancreatic cancer are inadequate. Therefore, researches on drugs that targeting the four high frequency mutations and their related pathways are important for precision treatment of pancreatic cancer. Compared with traditional pathological subtyping, transcriptomic subtyping uses mRNA (transcriptome) profiling to define intrinsic molecular subtypes, which can provide more valuable information for the implementation of precision therapy. The first transcriptome subtyping system for pancreatic cancer was proposed in 2011. After that, several studies have proposed new transcriptome subtyping systems. In general, according to the transcriptome profiling, pancreatic cancer can be defined as two subtypes: the classical and basal-like subtypes. The basal-like subtypes seem more resistant to chemotherapy and have a worse prognosis than the classical subtypes. In the future, advances in epigenetics and proteomics may provide new avenues for the clinical application of molecular subtyping.Personalized tumor model can screen potential effective drugs for cancer patients, which is an important approach for individual treatment. Compared with traditional tumor cell lines, patient-derived tumor xenograft (PDX) and organoid retain the molecular and biological characteristics of pancreatic cancer. Our center's practices on PDX and organoid research have confirmed that these two models have high success rate of cultivation, and organoid model can accurately predict patient response to chemotherapy. However, PDX and organoid have different tumor microenvironment from patients, and there is no standard for modeling methods and drug sensitivity test, which limits their application. Overall, the precision medicine approach for pancreatic cancer is still in the preliminary stage. The future researches should focus on finding new therapeutic targets for pancreatic cancer by means of multi-omics, establishing biobanks to accelerate preclinical researches of new drugs, and conducting prospective clinical trials to promote clinical translation of basic scientific achievements. In the future, the precision medicine will certainly offer more options for the treatment of pancreatic cancer.