One of the most frequently used parameters in water supply systems is the pipe breakage rate, which is often expressed by λ. Its definition can be deterministic or probabilistic, but both definitions lead to the same value under certain assumptions. This parameter can be defined per unit time only or per unit time and per unit length of the pipes (using a kilometre as the unit of measurement). The symbol λ is used in both cases and can occasionally generate confusion. In a water supply system, most failures occur because of pipe breakage, and only those cases are considered in this work. For a pipe, λ can be evaluated in terms of time (years) and length (km) and expressed by Λ. This paper aims to investigate the mechanical reliability of a water supply system with reference to pipeline breakages worldwide. A statistical approach is applied to a wide database of Λ values relative to certain variables (e.g., the diameter, material, age, pressure, and chemical characteristics of the water) for water supply systems worldwide (over 3,500 data points were collected from approximately 200 papers). The pipe length L and monitoring period T are often reported in the database. For each water supply system, it is necessary to apply a statistical weight to each value of Λ because the considered variables are notably different. The chosen weight is the product of L · T; the weight is evaluated using statistical analysis if L and/or T are unknown. Finally, a particular treatment is applied to the obtained weights to eliminate distortions in the Λ evaluation because of the different socioeconomic conditions of more or less developed countries. Four results are reported in this work: (1) the weighted average value of Λ in the world (Λm), (2) the average values of Λ in different continents (Λmc), (3) the average values of Λ in different countries (Λmn), and (4) the management correction factor (fCE).