Seventeen primary psychopaths, 17 secondary psychopaths, and 12 controls were studied. There were two administrations of the grid: (1) subjects were asked to construe as individuals and (2) subjects were asked how they thought people, in general, construed the situations. Results show that personal construct systems of primary psychopaths were characterized by extreme lopsidedness, and their sorting relationships showed great stability and consistency from their own individual grid to the people in general grid. However, the pattern of these sorting relationships was private and idiosyncratic. Secondary psychopaths showed little lopsidedness. The pattern of their sorting relationships was not as socially discrepant, and they exhibited some stability across administrations for their own elicited constructs. Normal control subjects were characterized by an intermediate degree of lopsidedness, stability in their elicited constructs, and high social agreement. These results suggest that primary psychopaths seemed to be making no distinctions between their own and people's, in general, conceptualizations. In addition, both psychopath groups showed a significant amount of misperception about people in general. 6 tables, 4 footnotes, and 16 references