Drinking water is a potential source of exposure to lead (Pb), which can pose risk to humans. The regulatory agencies often monitor Pb in water treatment plants (WTP) and/or water distribution systems (WDS). However, people are exposed to tap water inside the house while water may stay in the plumbing premise for several hours prior to reaching the tap. Depending on stagnation period and plumbing premise, concentrations of Pb in tap water can be significantly higher than the WDS leading to higher intake of Pb than the values from WDS or WTP. In this study, concentrations of Pb and water quality parameters were investigated in WDS, plumbing pipe (PP) and hot water tanks (HWT) for 7months. The samples were collected and analyzed on bi-weekly basis for 7 times a day. Several linear, non-linear and neural network models were developed for predicting Pb in PP and HWT. The models were validated using the additional data, which were not used for model development. The concentrations of Pb in PP and HWT were 1-1.17 and 1-1.21 times the Pb in WDS respectively. Concentrations of Pb were higher in summer than winter. The models showed moderate to excellent performance (R2=0.85-0.99) in predicting Pb in PP and HWT. The correlation coefficients (r) with the validation data were in the ranges of 0.76-0.90 and 0.97-0.99 for PP and HWT respectively. The models can be used for predicting Pb in tap water, which can assist to better protect the humans.