You're not using the 'parameters' argument to modelMatrix() correctly. The main point though is that there is no need to take special action to allow for different numbers of replicates. This is accommodated automatically in the course of an ordinary analysis. Gordon Dear Gordon, Sorry for being slow about this.... If in an extreme example you had three biological replicates A,B and C and for biological replicate A you had 12 replicates but for biological replicates B & C you only had 1 replicate Surely the model that is fitting the data for biological replicate A would be over emphasized relative to the others? The Experiment we are actually concerned with is descibed in the following targets list: (sorry its a bit complicated) Cy3 Cy5 pig1aA wtA pig2A pig1aA mybA pig2A tulipA mybA kal1A tulipA wtA kal1A pig1aA kal1A pig2A wtA mybA pig1aA tulipA pig2A kal1A mybA wtA tulipA tulipC pig1aC arpA wtC arpB mybB arpC tulipC wtB pig1aB wtC pig2C mybC wtC mybB tulipB pig1aC arpC pig1aB pig2B pig2B arpB pig2C mybC tulipB wtB Where A, B, and C designate the different biological replicates. So replicate A is covered by twice the number of slides as B or C. We are use to experiments with similar numbers of replicates within a loop design. However with this slightly scewed experiment we are worried that the data isn't accurately modelled. many thanks Jason -- -------------------------------- Jason Skelton Pathogen Microarrays Wellcome Trust Sanger Institute Hinxton Cambridge CB10 1SA Tel +44(0)1223 834244 Ext 7123 Fax +44(0)1223 494919