Both the well-established condition assessment and the newly developed assessment of a sewers structural substance are based on visual (CCTV)-inspection and defect coding, and thus a prone to errors and incomplete inspection data. In this study, we aim at understanding the sensitivity of structural substance assessment of sewers to several sources of uncertainties in the inspection procedure. We identified main uncertainty sources in the assessment procedure and propagated this uncertainty in the structural substance assessment of 80,000 sewer pipes. By comparing original inspection data with manipulated one, the effect of specific error types became quantifiable. It was observed that the use numerical assumptions for defect severity instead of performing a single case defect evaluation had a smaller impact on the substance assessment than on the condition assessment. However, the structural substance is more sensitive towards errors related to the extension of defects. While some errors can change the substance value / substance class of single reaches significantly, the mean impact is much lower. The substance assessment of a whole sewer network is thus quite robust against the here considered errors of inspection data.