Environmental epidemiological studies commonly quantify subjects' noise exposure level in their neighborhood. How this neighborhood is defined can vary across studies, leading to different approaches whose impacts on exposure levels remain unclear. This article examines the impact of the neighborhood's definition on environmental noise exposure estimates. L Aeq,24h exposures in the vicinity of 10,825 residential buildings were estimated using a high-definition noise map, built on a middle-sized French city. Various definitions of neighborhood (address point, façade, buffers, and official zoning) were used to produce different exposure estimates. Influence of urban environmental factors was analyzed using multilevel modeling. The results showed a significant increase of the exposure estimates (+3.9 dB) and a significant decrease of the variability, when the sample size of the considered neighborhood increased (P<0.01). The difference between the estimates from the 50-m-radius buffers and the 400-m-radius buffers ranged across buildings between –9.4 and +22.3 dB. This variation was influenced by urban environmental characteristics (P<0.01). Furthermore, the same approach was conducted individually considering assessments of exposure to road traffic noise railway noise and two atmospheric pollutants (NO2 and PM 10). The results highlight the need in further exposure and/or epidemiological studies to carefully consider neighborhood definition and environmental composition.