In 2000 and 2001 Canadians were shocked by water contamination events that took place in two provinces. In 2004 we undertook an internet-based survey across Canada that asked respondents to identify in percentage terms their total drinking water consumption according to one of three sources: tap water, bottled water, and home-filtered water (either some type of container or an in-tap filter device). In this paper we investigate the factors that influence these choices and whether choosing to either filter or purchase water is linked to perceptions of health concerns with respect to tap water. A series of one-way analysis of variance (ANOVA) tests suggest that the presence of children in a household and self-reported concern that tap water causes health problems lead to significantly greater consumption of bottled water or filtered water and significantly less tap water consumption. In order to examine these choices in a multivariate framework, we estimate a multinomial logit model. Factors yielding higher probabilities of a respondent being primarily a bottled water drinker (relative to the choice of tap water) include: higher income, unpleasant taste experiences with tap water, non-French-speaking, and being a male with children in one's household. Similar factors yield higher probabilities of a respondent being primarily a filtered tap water drinker. An important finding is that two key variables linking a person's health perceptions regarding tap water quality are significant factors leading to the choice of either filtered tap water or bottled water over tap water. They are: a variable showing the degree of health concerns a respondent has with respect to tap water and a second variable indicating whether the respondent believes bottled water is safer than tap water.