Introduction: Aripiprazole is hypothesized to have an effect on negative and cognitive symptoms in schizophrenia. Likewise, amisulpride is one of the only second-generation antipsychotics with which an effect on negative symptoms is reported. In the present study, we compare the effect of aripiprazole and amisulpride in initially antipsychotic-naïve patients with first-episode psychoses. Methods: Psychopathology and cognitive measures from two consecutive cohorts of antipsychotic-naïve first episode psychotic patients were obtained before and after 6 weeks of antipsychotic monotherapy with either aripiprazole or amisulpride. Matched healthy controls were included to account for retest effects on the cognitive measures. Analyses of variance (repeated-measures ANOVA) were performed to detect effect of time and possible cohort^*time interactions. Results: Longitudinal data was obtained from 47 and 48 patients treated for 6 weeks with amisulpride or aripiprazole, respectively. For the Wallwork negative symptom dimension, there was a cohort^*time interaction [F_{(1, 93)} = 4.29, p = 0.041] and a significant effect of time [F_{(1, 93)} = 6.03, p = 0.016], which was driven by an improvement in patients treated with aripiprazole [t₍₄₇₎ = 4.1, p < 0.001] and not observed in patients treated with amisulpride (p > 0.5). For the eight cognitive measures, no cohort^*time interaction was found and neither was cognitive improvement in any of the cohorts when accounting for retest effect. Conclusion: Patients treated with aripiprazole improved on negative symptoms, which was not the case for patients treated with amisulpride. This may point to a general effect of a partial D2 receptor agonist on negative symptoms in patients with first-episode psychoses. There was, however, no improvement in cognitive functions.