Parent-of-origin effect was examined in a series of 64 pedigrees with panic disorder (PD) under both the narrow and broad diagnostic models. The narrow diagnostic model defined the affected phenotype to include only the ''definite'' and ''probable'' forms of PD, whereas the broad diagnostic model included the entire PD symptomatology. The pattern of maternal vs. paternal transmission of disease was analyzed through a number of comparisons. These comparisons were performed first on all pedigrees and then on a subset of ''pure'' pedigrees including only strictly maternal transmission or strictly paternal transmission of PD. There were no significant differences in the proportion of offspring born to transmitting mothers vs. transmitting fathers under either diagnostic model or pedigree set. When the difference in the sex ratio among affected offspring from both transmission types was considered, only the ''pure'' pedigree sample under the broad diagnostic model yielded nominally (i.e., not corrected for multiple tests) significant results (P < .05). Also, the comparisons of cumulative lifetime risk for PD between offspring of transmitting mothers and fathers gave some nominally significant results; when affected and unaffected offspring were considered, significant results were observed under the narrow and broad diagnostic models, P < .0005 and P < .05, respectively. These findings must be considered provisional until confirmed by further study. Am.