Expectations for health-care quality affect how populations rate primary care, influencing tolerance of poor-quality care and demand for improvement, yet how expectations align with standards of practice remains poorly understood. We aimed to measure population expectations of primary care and their determinants within and across countries. Methods In this cross-sectional analysis, we used nationally representative data from the People’s Voice Survey collected between May 9, 2022, and Dec 21, 2023, in Ethiopia, Nigeria, Kenya, South Africa, Colombia, Peru, Mexico, Argentina (Province of Mendoza), Uruguay, Laos, India, China, South Korea, Romania, Greece, Italy, the UK, and the USA. Adults aged 18 years and older with at least one health-care visit in the past 12 months were included. Expectations were assessed using two anchoring vignettes, one depicting poor-quality primary care and one depicting adequate-quality primary care, administered via a mobile telephone survey, household interview, or nationally representative web panels. Respondents rated each vignette on a Likert scale: poor, fair, good, very good, or excellent. Low expectations were defined as rating the poor-quality care vignette as fair, good, very good, or excellent, and high expectations were defined as rating the adequate-quality care vignette as poor, fair, or good. Multivariable logistic regression models identified determinants of expectations, including covariates covering demographics, health status, care utilisation and experience, and system competence. Among 18 312 respondents (10 000 [54⋅6%] female and 8312 [45⋅4%] male; mean age 44 years [SD 17]), 2955 (16⋅1%) rated poor-quality primary care as fair or better and 11 787 (64⋅4%) rated adequate-quality primary care as less than very good. The proportion of respondents with low expectations of primary care ranged from 70 (6⋅3%) of 1112 in the UK to 530 (41⋅1%) of 1290 in South Korea. The proportion of respondents with high expectations of primary care ranged from 563 (40⋅1%) of 1404 in Nigeria to 494 (85⋅0%) of 581 in Italy. Older age, female gender, higher educational attainment, higher income, and use of private care (vs public care) were all associated with reduced odds of low expectations. Respondents with higher self-rated health had higher odds of low expectations (vs those with poor self-rated health). Activated patients (ie, those who were confident in raising concerns with their provider and that they were responsible for managing their own health) had lower odds of low expectations and lower odds of high expectations (vs non-activated patients). And positive perceptions of government management of COVID-19 were associated with high odds of low expectations and low odds of high expectations.