Cannabis use disorder occurs in up to 42% of subjects with schizophrenia. It has been associated with a lower age of onset of schizophrenia, more positive symptoms and a higher frequency of hospitalizations. A number of metabolomic-based studies have shown that levels of several amino acids and lipids are altered in plasma of schizophrenia patients. Similarly, different levels of various lipids, amino acids, sugars, proteins and other metabolites have been quantified in plasma of cannabis consumers, suggesting that consumption of cannabis does interfere in the regulation of further metabolic pathways. In the present study, we performed a non-targeted metabolomic approach in plasma samples of subjects with a diagnosis of schizophrenia, cannabis use disorder or dual diagnosis. The aim was to find potential biomarkers of the diseases and to gain a better understanding of the relation between cannabis abuse and the development of schizophrenia. The liquid chromatography-high resolution mass spectrometry outcomes allowed the annotation of 121 metabolites. The analysis between each group and the corresponding control samples showed a clearer clustering of the samples. The two-way ANOVA and linear analysis of the whole dataset showed that most of the analytes were differently present between the groups, but none of them seems to correlate to gender or age. It can be highlighted the presence of two endocannabinoids (palmitylethanolamide (PEA) and oleoylethanolamide (OEA)), nicotine and cotinine as biomarkers of tobacco consumption, and nor-9- carboxy-Δ9-THC as biomarker of the consumption of cannabis. Moreover, some medium chain acylcarnitines (C8/C10) share a common pattern in the three comparison groups whereas some N-acyl serines were also highlighted in some of the group comparisons. The use of this non-targeted metabolomic approach provided us a deep insight to evaluate the metabolomics profile and to uncover putative biomarkers in schizophrenia and cannabis use disorder.