Accurate determination of subsurface water levels is essential for the reliable quantification of hydraulic head gradients from which groundwater flow is inferred. Head measurements also underpin the quantification of hydrogeological properties based on aquifer pumping tests or tidal propagation analysis. While measuring a water level seems an easy task, the collection of water level time-series data is by no means trivial and there is a high potential for errors and misinterpretation. In this contribution we comprehensively analyse the individual measurements required to determine hydraulic heads and gradients: (1) geo-spatial positioning, (2) manual water depth, (3) automated pressure, and (4) spatial reference point for the head. For each component we determine the systematic and random errors based on our own data sets and field experience, as well as data from the literature. We assessed the effects of instrument and clock drift, variable density inside the piezometer, borehole inclination, well aging and instrument performance. By propagating the minimum achievable random measurement errors for horizontal and vertical hydraulic gradient calculations we evaluated the accuracy with which groundwater flow processes can be quantified using current best practice. The largest errors contributing to hydraulic head and gradient uncertainty originate from borehole inclination and manual water depth measurement, respectively. Our analysis demonstrates that resolving head gradients, which are smaller than 0.01 for boreholes that are closer than 10 m, requires extraordinary effort.