Over the past two decades Social Security Disability Insurance (SSDI) claims have been increasing as Workers Compensation (WC) claims have been falling.  Researchers interested in identifying a causal relationship between the two have focused on the impact of stricter WC laws on increasing SSDI applications and have found mixed results.  This paper builds on this literature by focusing on the impact of changes in Social Security disability’s effective replacement rates on the trends in both SSDI applications and reported workplace injuries.  Although the SSDI replacement rates for each of three defined earnings brackets have remained constant over this time period, the pivot point between earnings brackets increases at the same rate as average earnings. Growth in wages for workers in the lower half of the earnings distribution has been slower than the growth in mean earnings leading to increasing effective replacement rates for workers with low earnings.  After examining a disabled worker’s decision to apply for SSDI or WC, the paper presents both descriptive statistics and regressions with state level data to study the issue empirically.  Descriptive statistics show that the types of injuries and the types of workers who were more likely to report a workplace injury are now more likely to apply for SSDI.  At the state level, the number of recorded workplace injuries (used as a proxy for WC injuries) is regressed on the number of SSDI applications instrumented by the replacement rate of SSDI. The national trends of decreasing WC and increasing SSDI should be apparent at the state level if workers are substituting one program for the other.  Eligibility for WC and the level of benefits are determined at the state level.    While SSDI policy parameters are determined at the national level, the impact of these parameters will vary depending upon the distribution of earnings within a state and its evolution over time.  Variation in the level of state earnings creates variation in the relative generosity of SSDI for a given earnings quartile of workers.  Exploiting state variation in both WC and SSDI replacement rates makes it possible to identify the substitution of one set of benefits for the other. Results confirm the hypothesis that there appears to be substitution away from WC and towards SSDI.