Abstract Background Cardiac fibroblasts are considered to be highly heterogeneous mesenchymal cells after a cardiac injury. Understanding the heterogeneity of cardiac fibroblasts is of fundamental importance for fibroblast-targeted strategies to treat patients with heart disease. Advances in the single-cell RNA-sequencing (scRNA-Seq) technology have provided a possibility to elaborate the heterogeneity of cells in more depth. However, currently, most studies on scRNA-Seq of cardiac fibroblasts are performed using non-myocyte cells isolated at a certain point in time, which include a very complex, interdependent community of interstitial cells. scRNA-Seq based on lineage-tracing of cardiac fibroblasts is essential for a deeper understanding of cellular heterogeneity of specific cell type populations, which can reveal how a pre-existing cardiac fibroblast population responds to cardiac injury. Our previous studies demonstrated that a subset of cardiac fibroblasts can adopt the endothelial cell phenotype by undergoing mesenchymal–endothelial transition (MEndoT) and contributes to cardiac repair by revascularizing the damaged myocardium. Purpose To decipher which subgroup of the preexisting cardiac fibroblasts involved in MEndoT respond to cardiac injury. Methods and Results For lineage-tracing-based studies, Col1a2 Cre ERT: R26R td-Tomato transgenic mice were generated for reliable cardiac fibroblast tracking. 10X scRNA-Seq was performed to analyze the diversity of prelabeled cardiac fibroblasts after sham or transverse aortic constriction injury. Using the Seurat R package, the td-Tomato labeled cells were further clustered into 11 distinct populations. In addition, the cell trajectory assay shows both Cilp and Cd200 positive (Cilp+/Cd200+) cluster located in the early state and showing the phenotype of mesenchymal stem cells, indicating that this clusters have high diversity with regard to differentiation and directly contribute to angiogenesis. Besides, we isolated Cilp+/Cd200+ cardiac fibroblasts from wild-type mice by flow sorting and performed matrigel tube formation and acetylated low-density lipoprotein (Ac-LDL) uptake to measure the potential of MEndoT. Interestingly, compared with Cilp-/Cd200- cardiac fibroblasts, both matrigel tube formation and Ac-LDL uptake of the Cilp+/Cd200+ group were significantly increased (∼1.3 fold higher in matrigel tube formation, P<0.05;∼3.3 fold higher in Ac-LDL uptake, P<0.05). Conclusion Our findings indicated that Cilp+/Cd200+ is the main group of cardiac fibroblasts involved in MEndoT and plays a major role in the organization of the extracellular matrix and cell differentiation. This study provided new therapeutic strategies for cardiac hypertrophy by accurately regulating the cardiac fibroblasts.