This study investigated family dynamics and substance use disorders among the youth in selected rehabilitation centers in Nairobi County, Kenya. The study was guided by the family structure theory. The research design was mixed method embedded research design. Using Yamane (1967) formula, a sample of 172 clients was obtained and selected randomly to take part in quantitative study. Another 12 clients who had stayed in rehabilitation centers for the longest time were purposively selected in order to provide qualitative data. Similarly, 10 parents were conveniently selected in order to provide qualitative data. Therefore the total sample size comprised of 196 respondents. Quantitative data was collected using FACES-IV, AUDIT-10, and DAST-10 questionnaires while qualitative data was collected using interview guides. Cronbach Alpha technique was run to test the reliability of FACES-IV, AUDIT-10, and DAST-10. FACES-IV scale was found to have a reliability coefficient of .723, AUDIT-10 had reliability of .861while DAST10 scale had a reliability coefficient of .812. Quantitative Data analysis was done using descriptive statistics and inferential statistics while qualitative data was analysed thematically backed by narratives from respondents. The study had the following findings: Most families recorded unhealthy flexibility with majority of respondents coming from families which had chaotic family flexibility (Mean= 24. 4015; SD= 10.001) followed by Rigid flexibility (Mean= 17.4167; SD= 5.1244), there was a weak negative and significant correlation between balanced family flexibility and drug use disorder (r=-0.299; P= 0.001).