The criteria of most interest to regulators for measuring the environmental performance of a landfill cap is the quantity of water percolating from the base of the landfill cap into the buried waste. Phytocaps, also known as water balance covers or evapotranspiration covers, are nature based solutions comprised of a layer of loosely placed soil, planted with native trees and grasses, where the vegetation plays a crucial role in landfill cap hydraulic performance. In many countries, including in Australia, these landfill caps are required to demonstrate performance using lysimeters. This study sought to evaluate the hydrological performance of monolithic and capillary barrier phytocap test sections using lysimeters, and to assess the limitations inherent in this method of performance evaluation through comparison between lysimeters and unlined test sections. Four large test sections (20m x 20m area, 1.5 m deep) were designed, constructed, and instrumented at an operational commercial landfill site practicing leachate recirculation in Wollert Victoria, a temperate region of Australia. Landfill gas, temperature, soil moisture and plant growth differences were monitored over two years and three wet seasons. The establishing phytocap plant community was exposed to unmitigated landfill gas levels in the first 6 months, after which landfill gas extraction commenced. It was a further 9 months before landfill gas levels within the unlined phytocap test sections returned to normal soil-gas levels. Unlined test sections had higher volumetric moisture content (5% wetter) compared with the lysimeter test sections. Unlined test sections also experienced higher soil temperatures and landfill gas intrusion that impacted the plant community, with the presence of the lysimeter obstructing any interaction between the phytocap and the landfill. Whereas there was no distinction between the hydrological and ecological performance of the monolithic and capillary barrier designs when monitored with lysimeters, when the designs were monitored with the unlined test sections, the monolithic design clearly outperformed the capillary barrier phytocap design. This is likely because the capillary barrier layer acted as a gas distribution layer, an outcome of the design that could not be discerned from the lysimeter performance monitoring. These results suggest that comparative modelling of designs based on lysimeter data may not be meaningful for the deployment of the designs on a full-scale phytocap. Further, the limitations of lysimeters presented in this case study have significant implications for the interpretation of performance data obtained from studies based solely on lysimeter monitoring. Regulators, auditors, and designers need to consider: whether lysimeter trials are a necessary requirement in all contexts; whether to include unlined test sections to enable critical comparison; and, whether phytocapping is now a mature enough technology for performance to be demonstrated through the design process with due emphasis on soil and plants.