The heart is the first organ that forms and functions in the developing vertebrate embryo. It functions as a pump leading to blood circulation throughout the whole body. The cardiac conduction system is a group of specialized cardiomyocytes that initiate and propagate electrical impulses in order to coordinate heart contractions. The transcription factor Nkx2.5 is present early in the embryonic heart (at around E7.5) and is still expressed throughout the developing and adult heart. It also plays a role in maintaining the normal function of the cardiac conduction system. Conditional deletions of Nkx2.5 gene cause disturbance in heart development and cardiomyopathy in the mouse. Furthermore Nkx2.5 mutations have been found in patients with left ventricular non-compaction (LVNC). LVNC is a rare cardiomyopathy, characterized by hypertrabeculation and deep trabecular recesses in the left ventricle. This genetic disorder, associated with several mutations, exhibits clinical, morphological and functional heterogeneity. The main complications are heart failure, systemic embolic events and malignant arrhythmias. It is still unclear whether LVNC results from a defect occurring during cardiac development. One hypothesis to consider is that the severity of LVNC depends on which embryonic stage the arrest of myocardial compaction occurs. Our aim was to study the pathological evolution of LVNC by characterizing functional defects and identifying molecular mechanisms in mouse models with abnormal ventricular trabeculae development. To establish a LVNC mouse model, we generated specific Nkx2.5 conditional knockout by applying the Flox/loxP system inducible by tamoxifen injection to activate Cre recombination. This leads to the deletion of Nkx2.5 allele in atria and trabecular derived cardiomyocytes at embryonic stages when trabeculae arise (at around E10), or start to compact (at around E14), or at neonatal stages (after birth) when the heart is almost finish compaction step. To quantify the degree of non–compaction and the subendocardial and interstitial fibrosis, we carried out immunofluorescence analyses on left ventricular sections. Functional analyses show defects in ECG recordings with increased PR and QRS durations. This phenotype worsens with age with bundle branch blocks and a diminution in the ejection fraction at 6 months following deletion at E10.5/11.5 but not E13.5/14.5, consistent with a more pronounced phenotype after deletion at the earlier developmental stage. Microarray comparison of the transcriptomes of 6 month-old mice with different LVNC levels induced by Nkx2.5 deletion at the two timepoints, revealed differences in ventricular gene expression between the two groups, predominantly in pathways involved in calcium signaling, blood vessel development and immune or inflammatory processes. Deregulated expression of numerous genes playing a crucial role in cardiac cell activity or cardiomyocyte differentiation correlates with the different phenotypes observed in these mutant mice. To sum up, we were successful in generating several LVNC mouse models by the conditional deletion of Nkx2.5 transcription factor in atria and trabecular derived cardiomyocytes. These mouse models are suitable for studying LVNC pathology. We also confirmed the hypothesis that the severity of LVNC depends on stages when disturbances in the trabecular development occur. Hypertrabeculation, cardiac conduction defects, decreased ejection fraction, and existence of fibrosis are robustly observed following deletion at E10.5/11.5 meaning that the deletion at early stage of trabecular development causes the most severe pathological phenotype of LVNC. There had been just a few publications showing inflammation in LVNC heart, which could be a very good finding for future researches.