I am unsure if the package mi uses MICE because I can specify the number of iterations for the mi functions. Also on a side note: I have a large data set with 300+ covariates (some are have missing values). So I was wondering if I should use all the complete covariates for the imputation models or just the ones that are significant predictors (say at the five percent level). For example if I run a linear model for imputation not all the complete covariates are significant. Also if I run a glm with the missingness indicator as the response and the covariates as the explanatory variables, not all the covariates seem to be significant predictors. Thank you for your help. -- View this message in context: http://r.789695.n4.nabble.com/Imputation-using-package-mi-tp4639688.html Sent from the R help mailing list archive at Nabble.com.