The efficiency of a recently proposed novel global optimization method, energy landscape paving (ELP), is evaluated with regard to the problem of crystal structure determination from simulated X-ray diffraction data comprising integrated diffraction intensities. The new approach has been tested using the example of 9-(methylamino)-1H-phenalen-1-one 1,4-dioxan-2-y1 hydroperoxide solvate (C14H11NO.C4H8O4). The results indicate that, for this example, ELP outperforms standard techniques such as simulated annealing.