Patients diagnosed with schizophrenia often report a low quality of life (QoL). The purpose of this study was to investigate whether we could replicate a cross-sectional model by Alessandrini et al. (2016, n = 271) and whether this model predicts QoL later in life. This model showed strong associations between schizophrenia spectrum symptoms and depressive symptoms on QoL, but lacked follow-up assessment. This model was adapted in the current study and the robustness was investigated by using a longitudinal design in which the association between baseline variables (including IQ, depression, schizophrenia spectrum symptoms as well as social functioning) and QoL during 3-years of follow-up was investigated. We included patients with a non-affective psychotic disorder (n = 744) from a prospective naturalistic cohort-study. In the cross-sectional model, with good measure of fit, both depression as well as social functioning was associated with QoL (direct path coefficient -0.28 and 0.41, respectively). Additionally. the severity of schizophrenia spectrum symptoms was highly associated with social functioning (direct path coefficient -0.70). Importantly, the longitudinal model showed good measures of fit, which strengthens the validity of the initial model and highlights that depression prospectively affect QoL while schizophrenia spectrum symptoms prospectively influence QoL via social functioning. The negative, longitudinal impact of a depression on QoL highlights the need to focus on treatment of this co-morbidity. (C) 2019 Elsevier B.V. All rights reserved.