This paper discusses the issue of instrument proliferation in Difference and System Generalized Method of Moments (GMM) estimators, which are widely used in econometric analysis. The authors argue that these estimators, when implemented in popular software, often generate a large number of instruments, which can lead to overfitting endogenous variables and weaken the Hansen test of instrument validity. The paper reviews the literature on the effects of instrument proliferation and proposes simple techniques to control it, such as limiting the lag depth of instruments and collapsing instruments into smaller sets. The authors illustrate the dangers of instrument proliferation by replicating two early studies: Forbes (2000) on income inequality and Levine, Loayza, and Beck (2000) on financial sector development. Both studies found significant results driven by previously undetected endogeneity, highlighting the importance of addressing instrument proliferation to ensure the reliability of econometric findings.This paper discusses the issue of instrument proliferation in Difference and System Generalized Method of Moments (GMM) estimators, which are widely used in econometric analysis. The authors argue that these estimators, when implemented in popular software, often generate a large number of instruments, which can lead to overfitting endogenous variables and weaken the Hansen test of instrument validity. The paper reviews the literature on the effects of instrument proliferation and proposes simple techniques to control it, such as limiting the lag depth of instruments and collapsing instruments into smaller sets. The authors illustrate the dangers of instrument proliferation by replicating two early studies: Forbes (2000) on income inequality and Levine, Loayza, and Beck (2000) on financial sector development. Both studies found significant results driven by previously undetected endogeneity, highlighting the importance of addressing instrument proliferation to ensure the reliability of econometric findings.