Membrane proteins, believed to constitute between 20 and 30% of the proteome of most organisms and representing a considerable fraction of current drug targets, are notoriously difficult targets for experimental structure determination. At the New York Consortium on Membrane Protein Structure (NYCOMPS) we tackle membrane protein structure determination using a structural genomics approach. In this talk we discuss target selection and data analysis at NYCOMPS. Target selection's ultimate goal is to provide the experimental pipeline with 'well-behaved' proteins. We describe the several steps that we take in order to filter out problematic targets and to select the ones that may be more amenable to structure determination. Target selection is ultimately an iterative process, which uses analysis of the experimental pipeline's output to refine selection criteria. We present early data analysis at NYCOMPS and discuss future perspectives.