The issue of human-induced load and related mechanical performance has become one of the leading research topics in structural dynamics during the last decade. Although the concept of variability and uncertainty is well developed in structural dynamics disciplines such as wind, wave and earthquake engineering, most of the human-induced force models developed so far in structural engineering are deterministic, despite the intrinsic randomness of the crowd behaviour. The probabilistic models proposed in the last years have recognized two main sources of uncertainties, namely the structural system and the human-induced force. The pedestrian-related random variables usually considered in the cited force models are walking frequency, step length, free walking speed, single pedestrian force magnitude and body weight. According to the authors, another source of uncertainty should be considered, namely the one associated to the pedestrian traffic approaching and crossing the footbridge. This may find an analogy to what is usually done in wind engineering, where uncertainties related to the incoming wind are due to the inborn variability of the latter. Pedestrian traffic uncertainty is expected to deeply affect the crowd load on the structure: the magnitude of the overall force depends on the number of incoming pedestrians, while the load spatial distribution follows from the position of each pedestrian at the footbridge entrance and subsequently along the walkway. Coherently, in this study, the uncertainty evaluation includes the description of the undisturbed incoming traffic in terms of pedestrian density and hence introduces the variability of each pedestrian position. A two parts paper follows. The current one proposes a general procedure to collect in-situ data about the incoming crowd density and extract statistics. As such measurements are currently missing, a first attempt to obtain such statistics from data available in literature is proposed. The coefficient of variation of the incoming density is obtained and compared with the one of other pedestrian-related random variables.