Since 1990, almost 1.9 billion people have gained access to an improved sanitation facility, while 2.6 billion people still live without access to improved sanitation. In urban areas of low- and middle-income countries, the majority of household rely on onsite sanitation systems, producing huge amounts of faecal sludge (FS) that end in the environment every day. FS has to be managed in a proper way, integrating onsite system, storage, transport, treatment, disposal and reuse. In the past years, several FS treatment technologies have been developed and implemented, however partly due to the lack of knowledge on the FS characteristics, many of them have encountered failures. Reliable assessment of FS characteristics is essential in order to design adequate FS treatment facilities. However, FS is a highly variable waste and no methodology has been established to better understand this variability and the characteristics so far. This Master project, which takes place in the framework of the PURR project (Partnership for Urban Resource Recovery) that is a collaboration between the Sandec (Department of Water and Sanitation in Developing Countries, Switzerland), EPFL (Swiss Institute of Technology of Lausanne, Switzerland) and IESE (Institute of Environmental Sciences and Engineering of Hanoi, Vietnam), aims to answer this important gap of knowledge. To assess the influence of the local conditions on the characteristics of FS, fifty nine samples of household septic tank have been collected in Hanoi, the capital city of Vietnam. Fourteen factors reflecting local conditions and twenty eight parameters have been measured. QA/QC procedure together with the analysis of the organic compounds helped to determine the reliability of the data. Therefore, only reliable data was discussed, including seven factors (number of inhabitant, septic tank age, FS age, bio-additive, water addition in the septic tank, septic tank volume and number of trucks to empty the septic tank) and twenty parameters (temperature, pH, TS, TSS, VS, VSS, COD, carbohydrate, lipid, protein, VFA, TN, NH4+, TP, PO43-, total fibres, Ni, Pb, Fe and Zn). A degradation of the solids (TS, TSS), organic fraction (VS, VSS, COD, carbohydrate, protein, VFA) and nitrogen compounds (TN, NH4+) is mainly observed during the ten first years of storage. Addition of bio-additive potentially contributes to this process. The size of the septic tank seems to influence also the emptying frequency and then the degradation of the matter. Six sampling methods (S, HG, HT, B, M and D methods) have been tested over six samplings of household septic tank in Hanoi. First method with direct sampling inside the septic tank (S). Two methods collecting sludge from the manhole with different heights of collection (HG=15.4cm, HT=4cm). Three methods from the discharge valve of the truck (B=at beginning; M=at middle; D=at beginning, middle and end of the discharge with a volume ratio of 1:2:1 respectively). Sampling methods seems to not have influence on pH and temperature and soluble compounds concentrations. However, solids, organic content and nutrients concentrations are influenced. HG tends to under-estimate the concentrations when B and D could over-estimate it. HT and M seems to be the more accurate and reliable sampling methods of FS tested. Such knowledge on the FS variability could allow reducing the number of samples needed to obtain representativeness of FS. Indeed, collecting information related to the factors previously mentioned at the household level could work as predictor together with the characterization study to evaluate FS characteristics of a specific area. This allows savings in terms of time and budget. Moreover, the influence of the three factors (FS age, bio-additive addition, septic tank volume) on the FS characteristics, could constitute a basis for the development of policies and regulation. In order to reduce pollutant concentration in FS, appropriate emptying frequency and adequate storage tank design could be determined. The sampling methods HT and M would constitute low cost sampling methods, which would be simple to handle, fast, efficient and reliable to collect FS, and easily reproducible in different context. This is a real step forward for the improvement of knowledge about FS characteristics.