In Reply.—We fully agree with the comments by Drs Katz and McNamara regarding the various types of disease progressions in children infected with (HIV). Indeed, some patients with a "borderline" or an unusual progression might not perfectly fit into our description of a "bimodal population" of children according to the progression of the disease.1 We followed up a child, perinatally infected with HIV, who, at age 4 months, developed Pneumocystis carinii pneumonia. Three years later she was well, with no neurologic impairment or secondary infection. Her most recent CD4 cell count was over 1.5× 109/L, and the results of her in vitro lymphocyte proliferation tests became progressively normal, showing an association between the clinical and immunologic improvements. In most cases of HIV-infected children, there is a correlation between the severity of the symptoms and the impairment of the immune system or the intensity of the viral load. However,