Not all crystallographic structural investigations are amenable to a phasing solution by direct methods alone. Guideline procedures are outlined which are intended to help the evaluation of whether directmethods procedures may be expected to phase diffraction data for large molecular structures. This analysis is directed at three separate levels of inquiry:(1) How good are the primary data and can E values be derived to represent a point-atom structure. (2) How well do the data interact through phase relationships and may they be expected to produce a stable phasing solution. (3) What is the prognosis for finding recognizable solutions. Data are presented from the post-mortem analyses of a number of large, difficult-to-solve, structures to illustrate each of these points. Direct-methods practitioners are to be encouraged that crystal structures having more than 300 atoms per asymmetric unit may occasionally be determined utilizing present methodologies provided that an a priori prognosis for obtaining a solution is favorably high, adequate computational resources are available, and sufficient persistent effort is applied.