In an attempt to overcome the reflection overlap problem, which is the primary hindrance to structure determination from powder diffraction data, an experimental approach that exploits preferred orientation was developed some years ago. Now both the experimental setup and the data analysis procedure have been optimized, with the result that the quality of the extracted reflection intensities has been improved significantly and the synchrotron beamtime required for the data collection reduced. The one-dimensional Si microstrip detector Mythen II on the Materials Science beamline at the Swiss Light Source and new features in the data analysis softwareMAUD[Lutterotti, Matthies & Wenk (1999).IUCr Commission on Powder Diffraction Newsletter, No. 21, pp. 14–15] have made these improvements possible. The main changes in the experimental setup are (1) using an optimized set of sample orientations, (2) placing the detector such that both positive and negative 2θ angles are measured simultaneously, and (3) introducing an additional sample tilt angle, to measure data that cannot be accessed otherwise. On the data analysis side, the programMAUDis now used for both the determination of the orientation of the crystallites and the intensity extraction. The evaluation of data obtained from a textured zirconium phosphate pyridinium sample shows a significant improvement in the reliability of the structure factor amplitudes derived from overlapping reflections.