Objective To investigate the developmental trajectories of sleep quality in patients with schizophrenia and analyze the predictive factors of different trajectory categories, so as to provide a reference for developing targeted interventions to improve patients′sleep quality. Methods A total of 320 inpatients with schizophrenia were conveniently selected and assessed using the Pittsburgh Sleep Quality Index at discharge and at 1, 2, and 3 months post-discharge. Latent growth mixture modeling was used to analyze sleep quality trajectories, multivariate logistic regression identified influencing factors of different trajectories, and ROC curve analysis determined the predictive value. Results Totally 298 patients completed the entire follow-up. Sleep quality trajectories in schizophrenia patients were categorized into three classes: poor sleep-stable improvement type (34.90%), fair sleep-fluctuating improvement type (48.99%), and good sleep-rapid improvement type (16.11%). Age, alcohol history, use of medications that may cause drowsiness, disease course, Brief Psychiatric Rating Scale score, depressive symptoms, and current suicide risk were the influencing factors of sleep quality trajectory classes (all P<0.05). For the poor sleep-stable improvement type, the area under curve (AUC) was 0.913 (P<0.05), with a sensitivity of 0.865 and a specificity of 0.854; for the fair sleep-fluctuating improvement type, the AUC was 0.873 (P<0.05), with a sensitivity of 0.753 and a specificity of 0.896. Ten-fold cross-validation showed average AUC values of 0.868 and 0.837 for the two models, respectively. Conclusion Schizophrenia patients′ sleep quality trajectories are heterogeneous during the three months after discharge. Different sleep quality subtypes of schizophrenia patients should be identified in time, and targeted interventions based on influencing factors should be implemented to enhance their sleep quality.