Updated publication reference for PubMed record(s): 2566246, 25662467. Systems biological approaches to study the Arabidopsis thaliana circadian clock have mainly focused on transcriptomics while little is known about the proteome, and even less about post-translational modifications. Evidence has emerged that post-translational protein modifications, in particular phosphorylation, play an important role for the clock and its output. Phosphoproteomics is the method of choice for a large scale approach to gain more knowledge about rhythmic protein phosphorylation. Recent plant phosphoproteomics publications have identified several thousand phosphopeptides. However, the methods used in these studies are very labour-intensive and therefore not suitable to apply to a well-replicated circadian time series. To address this issue, we present and compare different strategies for sample preparation for phosphoproteomics that are compatible with large numbers of samples. Methods are compared regarding number of identifications, variability of quantitation and functional categorization. We focus on the type of detergent used for protein extraction as well as methods for its removal. We also test a simple two-fraction separation of the protein extract.