The study found that approximately one in eight Crown Courts in five circuits were awarding more custodial sentences to those pleading guilty than for those convicted at trial (reverse differential). Multiple regression analysis (MRA) of four administrative variables showed overall the not-guilty-plea rate was significant; for remanded defendants other administrative variables were relevant. Clearly, in terms of the MRA there were deviant localities from both positive-differential and reverse-differential courts. To examine this further, the analysis used descriptive statistics, which show that the MRA significance is produced by a minority of courts; the majority of positive-differential and reverse-differential courts have much in common administratively. An examination of the administrative characteristics of all these courts found a paradox; positive-differential and reverse-differential courts rarely share administrative characteristics, but they also vary greatly within each group. The study considered whether there are further advantages for defendants opting for trial in the reverse-differential courts in terms of their acquittal rates. On this evidence the study was unable to account fully for the two modes of criminal process found. 32 tables, 3 notes, and 4 references