Objective Cannabis is the most commonly used drug in the general population, but its prevalence of use remains higher among people suffering from severe mental disorders. Nevertheless, current cannabis research showed it to be deleterious on psychiatric symptoms, especially among patients with severe mental disorders. This present cross-sectional study aims to evaluate the impact of cannabis consumption on the psychiatric symptomatology of people with a serious mental disorder by controlling for the confounding variables of age, sex and concomitant alcohol or stimulant consumption. Method Secondary analyses were performed on data from 72 participants from a previous study. Their use of cannabis, alcohol and stimulants was measured using the Cannabis Use Problems Identification Test (CUPIT), the Alcohol Use Disorders Identification Test (AUDIT) and the frequency of use question from the Structured Clinical Interview for DSM-5-Clinician Version for Stimulant Use Disorders (SCID-5-CV-TLUS), respectively. Their psychiatric symptoms were measured using the five subscale model of the Positive and Negative Syndrome Scale (PANSS). Results Different linear explanatory models of PANSS symptoms were carried out using a combination of independent variables, i.e. age, sex, CUPIT, AUDIT and the question on consumption frequency of the SCID-5-CV-TLUS. The explanatory model of excitement symptoms is statistically significant (F = 4.629, p = 0.001) and it makes it possible to predict 20.4% of the variance of these symptoms (adjusted R2 = 0.204). In this model, CUPIT is the variable that most influences the model (ß = 0.381; p < 0.001). The explanatory model for positive symptoms is also statistically significant (F = 3.631, p = 0.006) and that makes it possible to predict 15.6% of the variance in these symptoms (adjusted R2 = 0.156). However, the CUPIT would not influence this model in a statistically significant way (ß = 0.125; p = 0.272), but the question on the frequency of consumption of the SCID-5-CV-TLUS would influence it (ß = 0.399; p = 0.001). In addition, the question on the frequency of consumption of the SCID-5-CV-TLUS also influences the explanatory model of excitement symptoms (ß = 0.273; p = 0.022). Conclusion Although further studies, ideally longitudinal, are needed to confirm the deleterious effect of cannabis on excitement symptoms, the present study reiterates the importance of screening and managing consumption habits of drugs, particularly cannabis, in people with serious mental disorders.