BACKGROUND: Around one in five people who use cannabis develop a cannabis use disorder (CUD) during their lifetime. Attentional bias toward substance-related stimuli is theorised to drive substance-seeking behaviour, contributing to the development of addiction and relapse. There is mixed evidence of cannabis-related attentional bias in individuals with CUD and its association with cannabis quantity, CUD severity, and confounders (e.g., alcohol, nicotine use). We aimed to examine cannabis attentional bias in non-treatment-seeking adults with a CUD (n = 74) versus controls (n = 30), and in relation to levels of cannabis and nicotine use, adjusting for alcohol use. METHODS: Attentional bias was measured using a visual probe task. To detect attentional bias differences between groups (CUD, controls), we ran linear mixed-effect models with reaction times (RTs) to cannabis versus neutral images as the outcome variable, adjusting for intra-individual variability and stimulus onset asynchrony (SOA, 200/500 ms). Within the CUD group only, we explore the association between attentional bias and cannabis quantity, craving, Cannabis Use Disorder Identification Test - Revised (CUDIT-R) scores, image valence ratings, and nicotine use. All models were adjusted for alcohol use. RESULTS: No significant effects were found for group, cannabis quantity, craving, CUDIT-R scores, or nicotine use on attentional bias toward cannabis cues. CONCLUSION: Findings suggest that non-treatment seekers with CUD do not exhibit attentional bias to cannabis cues. Replication in treatment-seeking samples is needed to determine whether attentional bias has clinical relevance, using objective measures (e.g., eye tracking).