The 1990s have been marked by several hydro-meteorological extremes, to suc
h an extent that it can be inquired if the riverside residents are exposed
to a risk, which is increasing with time. This perception of an increasing
risk could be due to a change in the frequency or intensity of the events,
to a change in the exploitation of resources, or even to the media coverage
of catastrophes. This article aims to study the progressive modification o
f the likelihood of flood event occurrence (maximum flow) in the Chateaugua
y River basin, based on available hydrometric data. The data has been studi
ed, using homogeneity and statistical control tests, to examine the possibi
lity of merging the data of two stations, in order to obtain a series, whic
h is sufficiently long to reveal the evolution of hydrological risk and to
study the change in the estimation of the 100-year return flood. The merged
series was then divided into subsets on which frequency analyses were perf
ormed using the linear moments method. Estimations of the 100-year return f
lood from different samples did not present significant differences. The bo
otstrap technique was then used in order to quantify the uncertainty due to
sampling. The results of the analysis show the absence of a general upward
or downward trend of the hydrological risk on the Chateauguay River.