Introduction Homeless individuals and migrants living in shelters—populations facing extreme vulnerability—have been disproportionately exposed to SARS-CoV-2 and have experienced higher rates of morbidity and mortality, primarily due to factors related to their living conditions (lack of access to healthcare, overcrowding). Vaccinating these individuals against COVID-19—a priority highlighted by numerous national and international recommendations—nevertheless faces several challenges, despite the efforts of many organizations on the ground and the introduction of the Health Pass in July 2021. Currently, no data exists on these populations’ access to the first dose of the vaccine, let alone on vaccination coverage. We therefore conducted a survey to estimate access to primary vaccination and vaccination coverage among this population and to identify and understand the factors associated with them. Methods We conducted a stratified cross-sectional survey using two-stage cluster random sampling among people in highly precarious living situations, defined by their place of residence, in the Île-de-France region and Marseille. The sample was stratified according to housing type, with six strata: 1) individuals housed in the Long-Term Housing Program, 2) in so-called “115” hotels, 3) in Migrant Workers’ Hostels, 4) individuals residing in Travelers’ Reception Areas, 5) homeless individuals living on the streets or in encampments, and 6) a subsample of a cohort of people in precarious situations in Marseille. Thus, between November 15 and December 22, 2021, 227 locations in the Île-de-France (IDF) region and in Marseille were surveyed, and 3,811 individual interviews were conducted in the participant’s language. Results In IDF, access to the first dose of the vaccine was 79.3% [95% CI: 76.0–82.6] in shelters, 70.4% [67.2–73.6] in “115” social hotels, 86.1% [83.3–88.7] in workers’ hostels, 41.3% [22.3–60.4] in permanent travel communities, and 44.3% [35.5–53.2] in shantytowns, informal squats, and on the streets. This coverage was 40.4% [30.8–50.0] among the homeless in Marseille. The rate of completion of the initial vaccination series (2 doses or equivalent) was 75.7% [72.2–79.3], 63.0% [59.5–66.5], 81.6% [78.7–84.5], 30.5% [14.0–46.9], 38.4% [30.4–46.5], and 32.4% [23.1–41.8]. People in highly vulnerable situations are less likely to be vaccinated than the general population across all age groups and also experienced a delay in access of approximately two months. Vaccination was primarily administered at vaccination centers open to the public (54.9% in total), while outreach programs accounted for 17.6% of those vaccinated. The reasons for not getting vaccinated were linked to refusal to get the vaccine (78% of reasons) rather than access barriers (22%), with the majority citing fear of side effects and a belief that the vaccine is ineffective. Furthermore, 24.2% stated that the Health Pass was the determining factor in their decision to get vaccinated. The univariate analysis identified numerous individual factors potentially associated with access to the first dose of the vaccine, including: gender, country of origin, age, immigration status, opinion on vaccination, fear of the vaccine, household composition, financial situation, certain support measures, sources of information on COVID, trust in authorities, health insurance coverage, and follow-up by a regular doctor. External structural factors explaining vaccination coverage include on-site vaccination organization, prior outreach, assistance with scheduling appointments, and transportation to a mobile vaccination clinic. Multivariate analysis shows that access to first-dose vaccination is correlated with age (OR >65 years vs. 18–25 = 2.4; 95% CI: 1.5–3.9) and is favored by being French-speaking (OR=1.3; CI: 1.0–1.6), holding (OR=2.4; CI: 1.8–3.1) or awaiting a valid residence permit (OR=2.0; CI: 1.4–2.7), having health insurance (OR=1.9; CI: 1.5–2.4), being under the care of a regular physician (OR=1.4; CI: 1.1–1.7), having a positive personal opinion about vaccination in general (OR=1.7; CI: 1.2–2.4), or having a social circle supportive of vaccination (OR=1.3; CI: 1.0–1.7). Seeking information about COVID-19 vaccination through one’s host promotes vaccination (OR=2.3; CI: 1.3–4.1), unlike using the internet and social media (OR=0.7; CI: 0.6–0.9) or the press (OR=0.7; CI: 0.5–0.9). In addition, needing the Health Pass (OR= 3.1; CI: 2.6–4.0), eating meals provided by one’s accommodation provider (OR= 1.8; CI: 1.3–2.4), having been hospitalized for COVID-19 (OR= 2.2; CI: 1.1–4.4), and trusting the authorities to manage the crisis (OR = 1.6; CI: 1.1–2.3) are also factors positively associated with access to the first dose of the vaccine. Conclusion Our study is the first in Europe to document access to vaccination among people in highly precarious situations. Our results show that despite well-documented overexposure to COVID-19, homeless and/or migrant individuals are less vaccinated than the general population, with vaccine access ranging from 40% in Marseille compared to 87% for the general population, and from 41% to 86% in the Île-de-France region compared to 91% for the general population. Above all, our results illustrate a gradient in vaccination coverage that can be considered parallel to the gradient of social integration: the more people have access to the mainstream system and/or are accompanied and supported by organizations, the more they obtain information from trusted individuals, and the higher the coverage. Our study also reveals that certain vaccination initiatives have been effective, such as vaccinations administered directly at accommodation sites or during food distributions. Furthermore, our results highlight the importance of the source of information regarding COVID-19 vaccination in the decision to get vaccinated or not. Our study has certain limitations, notably the difficulty in recruiting Travellers, the significant rate of replacement of sites and participants included, as well as the memory and social desirability biases inherent in this type of survey. Ultimately, barriers to vaccine access matter less than personal motivations in these populations. Mitigating these barriers, along with unconditional access to the vaccine, demonstrates that a proactive public health policy, supported by trusted mediators who know these individuals well, can yield positive results and should be expanded to encompass all aspects of healthcare access.