In this study of 197 community-recruited people who use unregulated opioids and cannabis, we found that self-reported cannabis use to manage unregulated opioid withdrawal symptoms was significantly associated with self-reported reductions in opioid use during periods of cannabis use. In sub-analyses stratified by pain, we observed that this relationship was modified by the presence of pain. Specifically, among participants experiencing moderate to severe pain, we observed a significant association between self-reported use of cannabis for opioid withdrawal and self-reported reductions in opioid use (p = < 0.001), whereas this relationship was not statistically significant among participants living with none or slight pain (p = 0.479). These findings contribute to the growing body of literature documenting how people at high risk of overdose may use cannabis as a strategy to manage their opioid use (Mok et al. 2023, 2021; Ceasar et al. 2021; Ganesh et al. 2024; Beaugard et al. 2024; Lake et al. 2020). Although prior research has examined the role of cannabis in mitigating opioid withdrawal symptoms, to our knowledge, this study is the first quantitative study to specifically investigate the relationship between intentional cannabis use for withdrawal management and changes in unregulated opioid use.

The observed association between self-reported cannabis use to manage unregulated opioid withdrawal symptoms and self-reported reductions in opioid use is biologically plausible as cellular receptor systems for opioids and cannabis are co-localized throughout the central nervous system and have been shown to produce analgesia via interaction (Wiese and Wilson-Poe 2018; Aquino et al. 2022; Scavone et al. 2013; Abrams et al. 2011). Pre-clinical, observational, and experimental studies have suggested that cannabis may modulate opioid withdrawal and alleviate symptoms (Aquino et al. 2022). A recent systematic review by De Aquino and colleagues synthesized observational and experimental studies evaluating the impacts of cannabis and THC use on symptoms of opioid withdrawal among people who use opioids (regulated or unregulated). The authors found mixed results in observational studies between cannabis exposure and alleviation of opioid withdrawal symptoms,experimental studies found more consistent mild-to-moderate effects of THC on alleviating withdrawal (Aquino et al. 2022). The authors propose that these discrepant findings between observational and experimental studies are likely a result of differences in participant characteristics, cannabis dose, and the type of opioids from which the participants are experiencing withdrawal (Aquino et al. 2022). The intersection of these factors likely moderates the association between cannabis exposure and opioid outcomes. Qualitative work also supports the findings we observed. For example, one qualitative study by Ganesh et al. among people who inject drugs in Los Angeles found that participants described using cannabis to reduce opioid use, including to manage symptoms of opioid withdrawal (Ganesh et al. 2024). Specifically, participants discuss using cannabis to manage the physical pain of withdrawal, and a number of participants also described that using cannabis to manage withdrawal also helped them reduce the frequency of opioid use (Ganesh et al. 2024). These findings are consistent with our observation that participants self-reporting cannabis use for opioid withdrawal were more likely to self-report reductions in opioid use.

Our findings may also be understood in the context of prior research supporting the potential use of cannabidiol (CBD), a non-intoxicating cannabinoid, to manage opioid withdrawal symptoms (Kudrich et al. 2021, 2024; Hurd 2017). For example, a double-blind randomized control trial conducted by Hurd et al. among drug-abstinent people living with opioid use disorder found that high doses of CBD (400 mg or 800 mg) versus placebo significantly reduced both drug cravings and anxiety following three days of administration, with effects persisting for up to one week after the final dose (Hurd et al. 2019). Anxiety is a key symptom of opioid withdrawal and has been found to increase the risk of relapse (Kudrich et al. 2021; Moradinazar et al. 2020). These findings may help contextualize the relationship observed in our study between self-reported cannabis use for withdrawal management and self-reported reductions in opioid use. Notably, doses of CBD administered in this trial likely substantially exceed the levels of CBD typically obtained through non-medical cannabis use, including cannabis products commonly used by participants in community settings from both regulated (e.g., cannabis stores) and unregulated (e.g., street-level dealers) outlets.

Our findings add to the growing observational literature describing real-world cannabis use behaviours among PWUD, including use with therapeutic or harm reduction intent, associated with beneficial outcomes (Mok et al. 2023, 2021; Lake et al. 2020; Fehr et al. 2025). For example, prior studies have found associations between cannabis use and reductions in opioid use and cravings (Reddon et al. 2023; Lake et al. 2019; Reiman et al. 2017), reductions in rates of injection drug use (Reddon et al. 2021, 2020, 2018), reduced odds of exposure to fentanyl (Socías et al. 2021; Hayashi et al. 2018), and improved retention in opioid agonist therapies (Socías et al. 2018; Raby et al. 2009). Our findings contribute to this growing body of evidence and demonstrate that intentional cannabis use to mitigate symptoms of unregulated opioid withdrawal may support reductions in opioid use, especially among people living with pain.

The findings in this study differ from those of Gorfinkel et al., who examined daily cannabis use and non-medical opioid use to assess potential substitution among people who use non-medical opioids (Gorfinkel et al. 2021). In that study, the odds of opioid use were approximately double on days when cannabis was also used, suggesting that cannabis did not function as a substitute for opioids in their study population (Gorfinkel et al. 2021). Although participant characteristics across studies were broadly similar, these divergent findings may reflect differences in study design and measurement. Specifically, Gorfinkel et al. assessed any cannabis use regardless of motivation and focused on same-day co-use, whereas the present study examined self-reported intentional cannabis use for opioid withdrawal management and self-reported changes in opioid use over broader periods. These approaches may capture distinct patterns of cannabis use, including general polysubstance use versus intentional therapeutic or harm reduction-oriented use.

As in previous studies, we detected an important role for chronic pain. There was a substantial prevalence of moderate to severe pain in our sample of PWUO (48.2%), consistent with prior estimates in similar populations (Voon et al. 2017, 2014; Lusted et al. 2013), which is likely not controlled through licit clinical means. Given the central role of chronic pain in unregulated opioid initiation, use, dependence, withdrawal, and relapse (Voon et al. 2018, 2015; Dahlman et al. 2017), the potential of cannabis to address pain-related symptoms may explain, at least in part, its role in facilitating reductions in unregulated opioid use. This is supported by prior research documenting associations between medical cannabis use and reductions in opioid medication doses among people receiving treatment for pain (Okusanya et al. 2020), as well as longitudinal research demonstrating a negative association between frequent cannabis use and unregulated opioid use among people who use drugs living with chronic pain (Lake et al. 2019).

Any possible benefits of cannabis use must be balanced by its potential risks, both acute (e.g., over-intoxication, accidents) and chronic (e.g., dependence), particularly among younger individuals or those with a personal or family history of psychosis, even if the potential nature and direction of effects of cannabinoids on psychosis remain unclear) (Walsh et al. 2017; Lake et al. 2020; Volkow et al. 2016; Ksir and Hart 2016). Similarly, people at high risk of fatal overdose through exposure to the unregulated drug supply might weigh their risk–benefit calculation differently. Importantly, cannabis is associated with markedly lower risks of morbidity and mortality compared to unregulated opioids, a conclusion supported by comparative drug harm assessments and epidemiological evidence (Lake et al. 2020; Nutt et al. 2010). Qualitative and ethnographic studies among PWUD further demonstrate that individuals’ perceptions of cannabis as a safer alternative align with these relative risk profiles (Paul et al. 2020; Ganesh et al. 2024; Beaugard et al. 2024; Valleriani et al. 2020). For example, in one study from the United States with interviews with people in recovery from drug dependence, all endorsed the idea that cannabis was a safer alternative to other drugs, pointing to its lack of contaminants, its legal status, and its inability to cause a fatal overdose (Beaugard et al. 2024).

Although consistent with the results of studies from our setting and others, these findings do not offer conclusive evidence supporting the use of cannabis for opioid withdrawal, which should be the subject of experimental study among humans. Nevertheless, even in the absence of evidence from controlled trials, policymakers could consider efforts to review and eliminate structural barriers to legal cannabis, including access and affordability, for PWUD using cannabis for therapeutic periods during the ongoing and unprecedented overdose crisis (Valleriani et al. 2020). All available evidence suggests that legal cannabis—from either the medical or non-medical cannabis regimes in Canada—remains largely inaccessible to PWUD, who continue to access cannabis through the unregulated market (Lake et al. 2020; Reddon et al. 2023). This is not only inequitable but also undermines the black letter public health goals of non-medical cannabis legalization in Canada’s Cannabis Act (Cannabis Act (S.C. 2018, c. 16) and the constitutional right of Canadians to access cannabis for medical purposes from a legal source (Parker 2000).

This study should be interpreted in the context of several limitations. First, participants in the VIDUS, ACCESS, and ARYS cohorts are not randomly selected, which may limit the generalizability of these findings to other settings. Second, given the cross-sectional study design, causality and the direction of associations cannot be established. Future research should incorporate more granular measures of opioid use, including quantity, to better capture the nuances of use patterns. Third, our reliance on self-reported substance use data introduces potential recall and social desirability bias. However, prior research has established that self-reported substance use patterns and behaviours among PWUD are valid and reliable when compared to biomarker assessments (Bharat et al. 2023; Darke 1998). In this study, participants were asked to reflect on perceived changes in opioid use during periods of cannabis use, which required participants to consider a connection between the frequency of use of two different substances. This may have introduced measurement error, including the potential for exaggerated or misattributed associations between cannabis use and opioid use. Alternative approaches, such as more frequent data collection (e.g., daily or weekly assessments), may provide more precise estimates of temporal relationships between cannabis use and opioid use. Finally, reductions in opioid use were assessed subjectively and in reference to periods of cannabis use. This measure reflects participants’ perceived reductions rather than objectively measuring changes and did not quantify the magnitude of reductions, nor or changes in actual opioid use frequency,as such, we are unable to determine whether reported reductions reflect clinically meaningful differences in opioid use. Moreover, the outcome measure did not include a parallel measure assessing perceived increases in opioid use during periods of cannabis use. As such, participants categorized as not reporting reductions may have experienced either no change or increases in opioid use, limiting our ability to characterize the full distribution of opioid use changes associated with cannabis use. Future studies should incorporate more granular and bidirectional measures of opioid use changes during periods of cannabis use.