The number of voltage dips, monitored in medium voltage networks, may somehow depend on certain characteristics of the network (network configuration, neutral management method, geographical area, length of overhead and cable lines, etc.), as well as environmental phenomena (lightning, presence of storm cells). The analysis was conducted by applying both correlation and multiple linear regression techniques on the time series of network and meteorological data. Durign the multi-year exercise, the voltage dips were detected by the monitoring system of the Italian medium voltage network QuEEN, while the characteristics of the network refer to the data provided in the past for this system. As regards meteorological factors, data on lightning density and the presence of storm cells were provided respectively by the Italian SIRF lightning localization system and by the STAF storm tracking system, based on the processing of radar reflectivity data. The percentage contributions of each network/environmental factor to the voltage dip statistics were assessed using time series correlation and/or multiple linear regression methods. The results of the data analysis confirm that the length of the overhead line has a significant influence on this disturbance, while the length of the cable lines has, as expected, only a negligible effect. The influence of lightning appears not to be as critical as expected and this also happens for the management of the network neutral, while the performance of voltage dips depends, in a non-negligible manner, on belonging to a geographical area. The influence of meteorological conditions is more evident if voltage dips are correlated with the occurrence of thunderstorm cells rather than with the density of lightning strikes on the ground.