The observed heterogeneity in negative symptom treatment response may be partly attributable to inadequate measurement tools or limitations in methods of analysis. Previous Item Response Theory models of the Positive and Negative Syndrome Scale (PANSS) have only examined samples of chronic patients and with mixed results. We examined the scalability of the negative subscale embedded in the PANSS and subsequently explored negative symptom trajectories across four weeks of amisulpride treatment. Data were derived from the OPTiMiSE trial comprising 446 patients with first-episode schizophrenia or schizophreniform disorder. Using the Rasch Model to examine psychometric properties of the PANSS negative subscale, we found that the composite score across items was not an adequate measure of negative symptom severity. Consequently, we chose an exploratory statistical approach involving Principal Component Analysis which yielded one significant component clustering into two significant symptom trajectories: 1) Subtle but constant decrease in negative symptom severity (N = 323; 72%), and 2) symptom instability across visits (N = 19; 4%). Explorative analytic methods as presented here may pave the way for more efficient and sensitive methods of analyzing negative symptom response in research and in clinical practice.