Consistency of diagnosti

li ne above which the diagnosis is said to be established. Data from the 1997 Australian Survey of Mental Health and Wellbeing were used for six internalizing disorders, and the levels of distress and disability associated with each diagnosis were investigated. Method: Mean distress (measured by the K-10) and disability (measured by the SF12-MCS) scores were identifi ed for people in the Survey who reported no physical or mental Mos that ex having being by the the wh in peo criteriodisorders. The distribution of distress and disability showed by people who met criteria for major depressive disorder, dysthymia, generalized anxiety disorder, social phobia, post-traumatic stress disorder and obsessive–compulsive disorder was plotted against the mean for well people, expecting that 90 % of people with these mental disorders would score as more distressed or disabled than this mean. Results: More than 90 % of people with dysthymia, major depressive disorder, generalized anxiety disorder or with post-traumatic stress disorder scored as more distressed or disabled than the mean for well people. A majority were severely distressed or disabled (2SD above the mean). This remained the case when the clinical signifi cance criteria were removed. In social phobia and in obsessive–compulsive disorders between 9 % and 26 % scored below the means for well people, that is, as neither distressed nor disabled, a fi gure that rose to 16–40 % when the clinical signifi cance criteria were removed. In neither case did a majority of cases score in the severe range. Conclusions: The diagnostic thresholds for social phobia and for obsessive–compulsive disorder are less stringent than that for the other disorders and require revision i