I found the UTF-8 problem still alive in group_by(). (Using the newest dev version. 0.4.3.9000) See below for reproducible code. Create a dataframe and make the colnames UTF-8. library(dplyr) df <-...