OBJECTIVES: The impact of addictive substance abuse (SA) on overall well-being, such as patient’s self-reported health status, may suffer from endogeneity bias. The purpose of this study is to estimate the impact of SA on patient’s health status using the instrumental variable (IV) approach. METHODS: The sample was extracted from the 2011 Substance Abuse and Mental Health Services Administration (SAMHSA) data for 36,347 adult US respondents who indicated SA status. Treating SA as exogenous, a single-equation probit model was used.  Treating SA as endogenous, bivariate probit models were employed using different IVs.  The outcome variable was the self-reported health status (1=better than good; 0=fair or poor).  Other exogenous variables controlled in the model were respondents’ age, gender, race, marital status, education level, employment status and family income.  IVs were respondents’ age at SA initiation, SA frequency and substance availability.  Exogeneity, over-identification and weak instrument tests were performed to ensure the validity and appropriateness of the IVs. RESULTS: About 50% of the respondents were between 18-25 years old.  Over 56% indicated SA and approximately 36% had initiated SA before age 18.  The single-equation probit suggested that SA was associated with at least a 2% decrease in overall health status (p<0.01), compared those who never had SA.  After adjusting for endogeneity bias, the negative relation between SA and health status increased by at least two-fold (p<0.01).  All tests indicated that the IVs used were valid and the results were significant. CONCLUSIONS: The findings suggest that estimating the impact of SA on health status should consider adjusting the endogeneity bias induced by individual characteristics and environmental factors such as age at SA initiation, frequency of SA use and substance availability in the community.  Further investigations on the effect of SA abstinence durations on health status should be explored to strengthen their associations.