In this paper, the stabilisation times for all 24 h of the day are analysed for 12 measurement stations located in the Spanish town of Malaga and throughout 5 years (2007-2011) of sampling environmental noise levels. For the results to be generalised to sound level measurements made in other streets or cities where there have been no long-term measurements, this study was developed for different road types and urban shapes. This distinction was made according to the types of roads indicated in other studies in which a statistically significant relationship was found between noise levels and the road type. The final objective of the study is to determine the capacity to estimate and approximate the real equivalent hourly noise level (L Aeq,1h) from the integration of a number of consecutive minutes (short-term measurements) less than 60 (L Aeq,1h ≈ L Aeq,T being T ≤ 60 min). Clearly, this strategy would save time and resources by making measurements of reduced duration. In summary and according to this analysis, a short-time measurement of 15 min is adequate to work with 90% confidence levels and errors of ±2 dB, with 80% confidence levels and errors of ±1 dB, and 50% confidence levels and errors of ±0.5 dB. However, it is necessary to consider the measurement hour period to achieve these levels of confidence due to the high variability throughout the day.