The crystal structures of two pharmaceutical compounds were solved from laboratory and high-resolution powder diffraction data collected at the ESRF Grenoble. The serious problems encountered during an ab initio crystal structure solution from powder data are described and discussed.