Dr. Rifkin anticipates our first response to his alternative interpretation of our results by noting in his letter that, by definition, primary negative symptoms do not come from positive symptoms. However, he claims that this assertion lacks empirical proof. In fact, this assertion has substantial empirical support from a number of studies conducted over the past 15 years by the University of Maryland group (1–5). Moreover, the results of our own study do not support the conclusion that olanzapine improved negative symptoms only when positive symptoms improved, even in the nondeficit group. In fact, we reported that positive symptom change, by itself, did not predict change in negative symptom score (p. 990). We believe that the most parsimonious explanation for our findings is that 1) olanzapine was effective in treating secondary negative symptoms because of its efficacy for positive symptoms and depressive symptoms, as well as its low propensity to cause extrapyramidal side effects, and 2) primary negative symptoms did not respond to treatment with olanzapine.