The muscular dystrophies are a group of genetically and clinically heterogenous disorders characterized by progressive muscle weakness and wasting. Animal models provide valuable insights into disease mechanisms. However, in light of the growing list of variants associated with human myopathies, patient-specific cell-based models could serve as complementary tools in disease modeling and therapy testing. Generation of cell-based models requires a cell source. Ideally, the cells would be easy to isolate and can be collected through a non-invasive method. We demonstrate that cells derived from urine samples can serve as source cells to be either directly reprogrammed into myotubes or reprogrammed first into induced pluripotent stem cells and then differentiated to cardiomyocytes. Generation of myotubes from patients with Duchenne muscular dystrophy and limb girdle muscular dystrophy 2C captured the patient-specific genetic alterations and recapitulated disease phenotypes, demonstrating the utility of urine cell-derived myotubes in disease modeling. One of the muscular dystrophies that has been challenging to model in mice is myotonic dystrophy (DM), as it arises from unstable repeat expansions in DMPK (DM1) or CNBP (DM2) gene. Transcription of these repeat expansions leads to formation of RNA foci that can sequester splicing regulators such as muscleblind-1 (MBNL1). This functional depletion of MBNL1 leads to missplicing of its target genes, shifting the splicing profile from adult to embryonic transcripts. We created cell-based skeletal and cardiac models from DM1 and DM2 patients in order to investigate the disease mechanisms in each tissue. DM1 myotubes and iPSC-derived cardiomyocytes displayed MBNL1 clusters and missplicing profiles, whereas DM2 cells exhibited MBNL1 distribution and missplicing patterns resembling those of control cells. In addition, iPSC-derived cardiomyocytes revealed calcium mishandling in DM1 and DM2. Combined, these findings suggest that DM2 may have additional mechanisms that lead to calcium mishandling in the absence of splicing aberrations. Transcriptional analysis of iPSC-derived cardiomyocytes implicated alterations in potassium channel expression as a potential mediator behind cardiac dysfunction in DM. These studies demonstrate that cell-based skeletal and cardiac muscle models can serve as a platform upon which disease mechanisms can be investigated. Additionally, the ability to generate patient-specific models could open up avenues for their use in testing gene-specific therapies.