Alterations in polyunsaturated fatty acids (PUFAs), including omega-3 and omega-6, have been implicated in the pathophysiology of psychotic disorders, but little is known about their associations with neuropsychological functioning. The present study includes 46 recent-onset psychosis patients who participated in a larger (n = 50) double blind, placebo-controlled randomized clinical trial comparing 16 weeks of treatment with either risperidone + fish oil (FO) (EPA 740 mg and DHA 400 mg daily) or risperidone + placebo and completed neuropsychological assessments at the baseline timepoint. We investigated the relationship between baseline omega-3 (i.e., eicosapentaenoic acid, EPA; docosapentaenoic acid, DPA and docosahexaenoic acid, DHA) and omega-6 (i.e., arachidonic acid, AA) PUFA with baseline MATRICS Consensus Cognitive Battery (MCCB) and Brief Psychiatric Rating Scale (BPRS) scores. Twenty-five patients had neuropsychological data available at 16 weeks following participation in the clinical trial, which included 12 patients assigned to risperidone + FO and 13 patients assigned to risperidone + placebo. At baseline both higher DHA and EPA correlated significantly with better social cognition after controlling for functioning on other neuropsychological domains, total BPRS score, AA level and substance use. Also, at baseline higher AA correlated significantly with hostility/uncooperativeness after controlling for DHA + EPA + DPA, overall neuropsychological functioning and substance use. Patients treated with risperidone + FO demonstrated a significant longitudinal increase in social cognition that was significantly higher at 16 weeks compared to patients treated with risperidone + placebo. DHA also correlated significantly with social cognition at the 16-week timepoint. This study provides novel evidence for a differential role of omega-3 vs. omega-6 PUFA in neuropsychological deficits and symptoms in recent-onset psychosis and its treatment.