Physico-chemical properties of residual waste are key factors for disposal standards. Reliable input data for treatment plants (MSWI) need to consider the heterogeneity and variability of waste composition. This requires thoroughly designed sampling schemes, a consistent separation of bulk waste into uniform fractions, and adequate methods for sample preparation and analysis. The present study aimed to provide up-to-date information on residual household waste properties, derive the underlying distribution patterns and delineate the contaminant loads associated with individual waste fractions. Data were gathered from composition analyses of 769 samples with a volume of 1,1 m3, each. Physico-chemical parameters as well as organic and inorganic contaminants were determined in 18 analytical substance groups (ASG) re-aggregated from 49 waste fractions . Extrapolation of analytical data to bulk waste was carried out by weighting the concentrations with the water content and mass fraction of ASGs using arithmetic means. On average, residual household waste was characterised by a water content of 37 mass-% and a lower heating value of 9.3 MJ/kg. Organic carbon, chlorine, and sulphur concentrations amounted to 22, 0.4, and 0.25 mass-%, respectively. Inorganic contaminants ranged from 0.1 to 320 mg/kg. Organic contaminants were dominated by polycyclic aromatic hydrocarbons with 2.7 mg/kg (Σ 16 PAH acc. EPA); chlorinated phenols (Σ mono- to pentachlorinated compounds) amounted to 170 μg/kg. Polychlorinated biphenyls (6 indicator congeners) totalled 70 μg/kg and polychlorinated dioxins/furans were determined as 5.5 ng I-TEQ/kg (NATO CCMS). In addition to calculations based on arithmetic means, Monte Carlo methods were employed to account for the spreading of data and waste variability. While the prior gives absolute values of waste properties this procedure yields frequency distributions. For inorganic and organic contaminants these were distinctly skewed reflecting the spreading of analytical results and lognormal mass distributions of ASG. Thus, especially for the use as design data, residual household waste characterisation by arithmetic means may be insufficient. Results of Monte Carlo methods give a more realistic view: Since frequency distributions can be expressed as percentiles of probability functions the significance of results is substantially increased.