The Sustainable Development Goals recognize that the availability and quality of improved water sources affect how households use and benefit from these sources. Although unreliability in piped water supplies in low- and middle-income countries (LMICs) has been described, few studies have assessed household coping strategies in response to unreliable water supplies and associated health outcomes. We characterized unreliability in the piped water supply of the town of Borbón, Ecuador over the twelve years following a major upgrade, as well as household coping strategies and associations with diarrhea. We examined trends in primary and secondary drinking water sources, water storage, and water treatment using longitudinal data collected from 2005 to 2012. In 2017, a follow-up survey was administered (N = 202) and a subset of 84 household water samples were tested for chlorine residual levels and microbial contamination. From 2005 to 2017, access to a household water connection increased from 19.4% to 90.3%. However, reliability decreased over time, as in the latter half of 2009, households had access to piped water 79% of the time, compared to 63% by 2017. Piped water samples were highly contaminated with total coliforms (100% of samples) and Escherichia coli (89% of samples). From 2005 to 2017, households less likely to report drinking water treatment (50.6%-5.0%). And from 2009 to 2017, bottled water was increasingly consumed as the primary drinking water source (18.8%-62.4%). From 2005 to 2012, having a household connection was not statistically significantly associated with diarrhea case status (OR: 0.86 95%CI: 0.53, 1.39). Neither household water treatment nor bottled water consumption were negatively associated with diarrhea. Increased water storage was associated with diarrhea (OR: 1.33 per 10L of water stored, 95%CI: 1.05, 1.69). Household water treatment, and consumption of purchased bottled water, two coping strategies that households may have undertaken in response to an unreliable water supply, were not associated with a reduced likelihood of diarrhea. These data suggest a need to understand how impoverished rural households in LMICs respond to unreliable water supplies, and to develop heath messaging appropriate for this context.