An automatic fire sprinkler piping system is an essential non-structural element which can have major impacts on both the fire safety and building functionality (e.g. flooding damage) if damaged during a seismic event. Generally, an intricate piping network is required to feed individual sprinklers which are spread across the plenum space. Depending on their function, the pipes in a network can be of varying diameters and lengths, and are usually attached to the floor at different depths relative to each other. The varying geometric and configuration details of the different segments of piping implies different dynamic characteristics, and thus a piping network subjected to a floor acceleration will have varying local seismic demands. In practice, this elaborate piping network is braced against seismic demands using proprietary braces that are installed following a set of empirical design provisions (primarily spacing) given in the New Zealand Standard (NZS) 4541: Automatic Fire Sprinkler Systems. Relying on these provisions, the NZS 4541 requires the sprinkler system to be operational at the ultimate limit state (ULS) earthquake loading. However, the description of typical piping installation and design practices implies that bracing a random piping system using empirical design provisions, without any regard to the actual seismic demand, leads to a system whose expected global performance cannot be reliably defined. Precisely, it is not known whether the design provisions are robust enough to ensure that none of the individual segments undergo damage during design intensity shaking, which could compromise the functionality of the whole system. To address this problem, shake table tests on sprinkler piping systems with multiple variations that are representative of actual practices are being conducted. The seismic provisions of NZS 4541 are followed when determining specimen brace locations and types. The piping specimen is mounted on a braced steel frame, and consists of a distribution pipe connected to a branch pipe through a riser nipple. The branch pipe is further connected along its length to three armover pipes of different lengths. The hanger rods and braces are anchored into individual concrete blocks attached to the test frame. The pipes are filled with water to monitor leakage of the piping connections during shaking. The set of input table motions include recorded floor acceleration response histories of an instrumented building in New Zealand. This paper describes the test setup and piping configuration. An initial set of results from the ongoing experimental program are presented and their implications for the design of sprinkler piping systems in New Zealand are discussed.