The eddy covariance technique was deployed to measure carbon dioxide (CO₂) fluxes over a two-year period in a suburban area of Beijing, China. The contributions of anthropogenic and biogenic sources and sinks to the CO₂ budget were quantified using a combination of statistical data and empirical models. Long-term weekly CO₂ fluxes exhibited pronounced seasonality, with daily mean values ranging from -0.9 μmol m⁻² s⁻¹ in summer to 11.4 μmol m⁻² s⁻¹ in winter. Diurnal bimodal patterns were detected across all seasons, with the morning peak timing varied seasonally attributed to discrepancies in the development of the turbulent boundary layer, whereas the evening peak coinciding with the traffic rush hour. From spring to autumn, daytime fluctuations in CO₂ fluxes were predominantly driven by photosynthetic processes, leading to net CO₂ uptake by the ecosystem. Differences between weekday and weekend/holiday flux patterns were more evident in autumn and winter, likely due to behavioural adaptations to colder temperatures, such as delayed outdoor activities and earlier returns home. Annually, the study area functioned as a net CO₂ source, emitting 5.51 kg m-2 year-1 on average. Traffic emissions constituted the largest share at 3.10 kg m-2 year-1, followed by building emissions of 2.48 kg m⁻² year⁻¹. Human metabolism is estimated to contribute 12.3%. Although photosynthetic uptake offset 58.4% of the gross annual CO₂ flux, biogenic respiration counteracted this removal by contributing 44.8% to the total emissions. Consequently, net ecosystem CO₂ uptake accounted for 13.6% of the annual CO₂ emissions.