The aim of the paper is to evaluate the relative importance of the set of socioeconomic characteristics of population on collective decision on COVID-19 vaccine acceptance. We apply cross-section OLS methods to the municipal-level non-survey data for 145 municipalities in Serbia, on the COVID-19 vaccination rate and socioeconomic characteristics of the population, to evaluate the determinants of cross-municipal variation in vaccine uptake decision. Using the estimated coefficients from the OLS regressions, we apply the standardized beta method to evaluate the relative importance of each factor. Vaccine acceptance in municipalities rises with the average level of education (especially in the female population), age and employment, while being negatively linked to religiosity of people and the proportion of rural population. We also find some evidence on the positive impact of the overall trust in government. Education level has the single largest impact, shaping around 37% of (explained) variation in the vaccination rate across municipalities, a rise in the proportion of people with higher degree by 1% being associated with increase in vaccination rate by 0.36%. Age of population explains 21%, urban-rural structure 13% and religiosity 11% of variation in vaccine acceptance, while employment status and trust in government each explain around 9% of variation in vaccine uptake across municipalities. Effective vaccination promotion strategy should be focused on younger, less-educated, unemployed cohorts, as well as on rural areas and should involve representatives of mainstream religions. Fostering education and strengthening trust in government are some of the key structural factors that may promote efficient collective behaviour in this respect.