Cannabis sativa Linne, an annual plant native to Central Asia, has long been cultivated worldwide. In Japan, its fibers and seeds have been traditionally used for purposes such as making sacred shrine ropes and as ingredients in spice blends such as shichimi togarashi [1]. Cannabis contains more than 160 bioactive compounds collectively known as cannabinoids [2]. Cannabinoids exhibit various pharmacological effects, including analgesic [3], anticonvulsant [4], and anti‐inflammatory [5] actions, and have been recorded as medicinal agents since ancient times [6]. Among these cannabinoids, Δ9‐tetrahydrocannabinol (THC) is the most abundant and is known for its characteristic psychoactive effects [7]. Due to these properties, cannabis was classified as a Schedule IV substance under the 1961 Single Convention on Narcotic Drugs and has been regulated globally as a highly addictive drug [8].

The assessment of cannabis‐related health risks has changed over time depending on sociocultural and historical contexts. In the early 20th century, international regulation became stricter, influenced by antimarijuana campaigns, especially in the United States, being driven more by political and social factors than by scientific evidence [9, 10]. Since the 1990s, research has revealed the existence of endogenous cannabinoids and their receptors and related enzymes in the human body, known collectively as the endocannabinoid system (ECS) [11]. This discovery has positioned the ECS as a target for the treatment of various disorders. Consequently, countries such as the United States, Canada, and parts of Europe have expanded the legalization of medical cannabis, prompting a reevaluation of cannabis' risks and benefits based on scientific data. Recent studies have suggested that cannabis is somewhat less harmful than legal substances such as alcohol and tobacco [12].

Although cannabis is often regarded as less harmful than other illicit substances, concerns remain about its potential role in facilitating progression to the use of more dangerous drugs. This concern is encapsulated in the so‐called “gateway drug hypothesis,” which originated from epidemiological and longitudinal studies conducted primarily in the United States since the 1960s. The hypothesis posits that individuals who begin using cannabis are at greater risk of subsequently engaging in the use of harder drugs, such as cocaine or heroin [13, 14]. Numerous studies supporting this hypothesis have consistently shown that cannabis users report higher rates of subsequent illicit drug use compared to nonusers, with earlier initiation of cannabis use further increasing this likelihood [13, 14]. However, the majority of these studies are correlational in nature and fall short of establishing a direct causal relationship. They merely indicate that a proportion of cannabis users also engage in other drug use, without sufficiently accounting for key confounding factors such as social, psychological, and environmental influences, or individual personality traits.

Animal studies have also suggested that exposure to cannabinoids may alter the brain's reward system, potentially enhancing sensitivity or tolerance to other substances [15, 16, 17]. Nonetheless, the extent to which these findings can be generalized to humans remains contentious, given the complexity of human behavior and environmental factors.

Alternative viewpoints suggest that individuals who initiate cannabis use at a young age may already possess underlying vulnerabilities, such as risk‐taking personality traits, adverse family environments, or social networks where polysubstance use is common [18, 19, 20]. These factors may independently contribute to the observed association between cannabis use and subsequent engagement in other drug use.

Despite these limitations, the gateway drug hypothesis continues to shape drug policy worldwide, including in Japan. The Japanese government has cited this hypothesis as a key rationale for reinforcing cannabis regulations. Prior to the 2023 amendment of the Cannabis Control Act, which introduced criminal penalties for cannabis use [21], the Ministry of Health, Labour and Welfare (MHLW) presented materials referencing this hypothesis during expert panel discussions [22]. These materials drew on findings from a Ministry of Justice report (White Paper on Crime, 2020) [23], which showed that approximately half of stimulant offenders had a history of prior cannabis use, with many initiating cannabis use before the age of 20.

Importantly, this government report focused exclusively on a population of incarcerated individuals detained under Japan's Stimulants Control Law, limiting the generalizability of the findings to broader drug‐using populations. Moreover, while cannabis was reported as the most common “first drug” among offenders under 30 years old (42.6%), organic solvents were the most common first substance among offenders aged 30–49, and methamphetamine was most common in those aged 50 and older. Despite this, the focus was primarily placed on the under‐30 group, while the trends observed in older cohorts were largely disregarded.

Another critical limitation is that tobacco and alcohol—known common precursors to cannabis use—were excluded from the survey's list of substances, thereby preventing a full understanding of the true initiation patterns. While these findings have been cited to support the narrative of cannabis as a gateway drug, the methodological constraints and selective interpretation of data raise questions regarding the robustness of this claim in the Japanese context.

Against this backdrop, we conducted a secondary analysis of nationwide online survey data from cannabis users in Japan, aiming to critically reexamine the validity of the gateway drug hypothesis in this context. Specifically, we sought to clarify the patterns and sequences of cannabis and other drug use among community‐based users.