The human health impacts of non-tuberculous mycobacteria (NTM) in home environments remain poorly understood. To better understand the association between NTM colonization of POU sources in homes and MAC pulmonary disease (MAC PD), we continued a case-control study of Washington and Oregon residents who have been diagnosed with MAC PD and population controls matched by age, gender, and geography. Environmental samples were collected from bathroom faucets, kitchen faucets, shower aerosols, indoor soil, and outdoor soil. Mycobacterial load in these samples was quantified by using bacteriological culture and PCR. NTM isolates obtained from 30 to 40 matched sets of case and control homes (depending on POU source) were quantitatively compared in three conditional logistic regression models with varying levels of control for confounding. An age-adjusted conditional logistic regression analysis of NTM isolates from shower aerosols shows that the homes of cases had higher odds (odds ratio, 3.67, 95% confidence interval, 1.16-11.53) of having NTM colonization compared to the homes of their matched controls. Associations for the other POU sources remain uncertain; the odds ratios estimated for other POU sources were consistently greater than one, but with wide confidence intervals, limiting the conclusions that can be drawn without further data. To better understand what house and homeowner characteristics may affect the colonization of NTM in homes in water POU sources, we also conducted an exploratory analysis utilizing the previous case-control study as a cross-sectional study using various logistic and linear regressions. Associations for various predictor variables relating to house or homeowner characteristics to NTM colonization remain uncertain; however, the large effect estimates and wide confidence intervals suggest that more data in a future study would improve precision and potentially be able to better detect associations. To our knowledge, this is the first etiologic epidemiological study on the association between NTM colonization of a specific household site and human disease. Previous studies were unable to connect the exposure site to pulmonary disease. The results implicate shower aerosols as a potentially significant source of NTM exposure in homes. While we were unable to detect associations between house and homeowner characteristics and NTM colonization in our exploratory analysis, these characteristics should be included in future study designs with the help of a qualified technical expert.