Reports an error in “Why and when hierarchy impacts team effectiveness: A meta-analytic integration” by Lindred L. Greer, Bart A. de Jong, Maartje E. Schouten and Jennifer E. Dannals (Journal of Applied Psychology, 2018[Jun], Vol 103[6], 591-613). In this article, the hierarchical form value for the study by Cantimur et al. (2015a) was incorrectly coded as representing ‘acyclicity’ rather than ‘steepness’. Updating the coding for this study means that there is just one acyclicity study in our dataset (Bunderson et al., 2016) rather than two, and resulted in some changes to the coefficients in our moderator model presented in Table 3. The corrected Table 3 is presented in the erratum. The only substantive change to our conclusions is that Hypothesis 4a, regarding the effects of membership instability, is no longer supported at p < .05. We also note a minor change to how Hypothesis 5b is tested–we can no longer test for the effects of acyclicity at the study level of analysis. However, we can still test this hypothesis using the 'shifting unit of analysis' approach. These analyses revealed that the population coefficient for acyclical hierarchies was unlikely to be different in magnitude from that of steep hierarchies or centralized hierarchies. Our original conclusion and interpretation–that Hypothesis 5b is not supported, but this interpretation is limited due to low sample size–remain intact. See erratum for full description. (The following abstract of the original article appeared in record