Lay abstract In previous schizophrenia (SZ) studies, many long noncoding RNAs (lncRNAs) have been shown to be expressed differently in patients with SZ compared with healthy controls. This suggests that lncRNAs are involved in the development of SZ. However, there is limited information about the relationship between lncRNAs and symptoms in SZ. The diagnostic and predictive value of nine lncRNAs was studied in 48 patients with SZ before and after antipsychotic treatment. The results indicated some of these lncRNAs may be useful in predicting outcomes of antipsychotic treatments in patients with SZ. This is important for future research developing biomarkers for the diagnosis, treatment and prognosis of SZ. Aim: The aim of this study was to identify the long noncoding RNAs (lncRNAs) associated with schizophrenia (SZ) and the relationships among their expression, antipsychotic efficacy and SZ severity. Method: The diagnostic and predictive value of nine lncRNAs, Gomafu, DISC2, PSZA11, AK096174, AK123097, DB340248, uc011dma.1, ENST00000509804-1 and ENST00000509804-2, was investigated in 48 patients with SZ before and after antipsychotic treatment. Results: Gomafu, AK096174, AK123097, DB340248, uc011dma.1, ENST00000509804-1 and ENST00000509804-2 were individually and collectively associated with, and predictive of, SZ pathogenesis. Moreover, increased expression of plasma AK123097, uc011dma.1 and ENST00000509804-1 levels was reversed after 12 weeks of antipsychotic treatment, which was associated with SZ severity. Conclusion: Seven lncRNAs serve as novel biomarkers for SZ diagnosis and prognosis and three lncRNAs are potential therapeutic targets.