Structural behavior of flexible plastic manholes is very different from rigid concrete products. Although this is state-of-the-art knowledge, no international or EN standard exists on structural design of these products. Therefore, traditional soil models and soil pressure coefficients are usually applied if calculations for design checks are required. These models are based on rigid wall structures and pipe design methods and do not adequately account for the flexibility and soil interaction of the manholes in the subsurface 3D-space. Recently a working group was established in Germany to develop a design standard for concrete and plastic manholes. In order to calibrate the design equations for the flexible manholes on experimental data and field experience, an extensive in-situ testing of a PP manhole (DN 1000) in a sandbox was conducted. The comparison of deflection readings and calculation result reveals a great difference. Evidently, the assumption of horizontal soil pressure on plastic manholes is wrong. Only soil pressure coefficients of less than 0.1 give realistic results. This is significantly less than the coefficient for active soil pressure, which is widely used today. Short Summary: The paper presents results from sandbox testing of plastic manholes. It focuses on horizontal deflections resulting from wheel loads at the surface. Explanation is given on why most part of vertical point loads adjacent to flexible manholes is not acting on these products. Theoretical background for this specific behaviour is the load distribution of vertical soil pressure in the subsurface 3D-space. Based on these findings, a range for realistic soil pressure coefficients for actions on plastic manholes is presented.