The dates and places of 105 suspected or proven murders committed by Dr H. Shipman of Hyde in Greater Manchester between 1984 and 1998 were subjected to epidemiological analysis. These interim data were assembled through a variety of media sources and were based upon court proceedings and police investigations. The analysis revealed a significant excess of case pairs separated jointly by distances <0.6 km and by times <7 days. These sequences were also associated with purely geographical groupings within spatial diameters <0.2 km, but with longer time intervals. This was confirmed by an extended analysis of case triplets occurring within short times and distances. They showed a significant excess. These clusters were probably generated by a combination of psychopathic obsession, of local opportunity, and of caution; and in similar circumstances could occur again. The findings demand the regular display of mortality data in a format permitting the intuitive recognition of similar phenomena elsewhere.