Dear All, I have a quick question regarding using groupedData in the nlme package. Can I add more than one predictor in the groupedData formula? I can't seem to get this to work and all the examples in Pinheiro and Bates show just one predictor in the groupedData formula. Can I use a more complex model when I have only one predictor in the groupedData formula? Currently I am using the following formula: grouped<-groupedData(y ~ 1 | ID, data = dataset) I would like to run the following models: fixed<-lme(y ~ -1 + x1 + x3 + x3, data = grouped) randome<-lme(y ~ -1 + x1 + x2 + x3, random = list(~x1|ID, ~x2|ID, ~x3|ID), data=grouped) Thanks, Alex