Identifiability is an important aspect of model development. In this work, using a simple one compartment population pharmacokinetic model, we show that identifiability of the variances of the random effects parameters are affected by the parameterisation of the fixed effects parameters.