The consumption of goods and services generates material flows causing environmental impacts. In cases where the environmental impacts from consumption has been calculated for a single municipality, national economic flows are generally transformed to material flows and subsequently downscaled to municipal level. This method is called economic input-output LCA and does to a very limited extent account for local differences in the consumption patterns. In this pilot study it was therefore investigated how material flow analysis (MFA) can be combined with life cycle analysis (LCA) in order to develop a more precise method for calculating the environmental impact from the consumption of goods and services. The scope of the pilot study includes food and food waste flows in the municipality of Lerum. Data regarding food flows were calculated using the MFA model Urban Metabolism Analyst (UMAn). UMAn enables calculations of product and material flows on municipal level regarding e.g. import, export, extraction, production, consumption and accumulation of material flows. Data regarding food waste flows were collected from the municipality of Lerum. The MFA data for food and food waste were subsequently combined with LCA data from the IVL Swedish Environmental Research Institute’s LCA data bank. The environmental impact categories addressed in the pilot study are acidification, formation of ground level ozone, greenhouse effect, eutrophication and primary energy use. The calculated environmental impact from food consumption is presented in section 3. According to the calculations, food consumption in the municipality of Lerum caused greenhouse gas emissions equivalent to 33 000 tons carbon dioxide equivalents in 2011. This corresponds to almost 0.9 tones carbon dioxide equivalents per capita. The food category meat (animal fat, pork, chicken and bovine meat) was identified as an important category to influence in order to reduce the environmental impacts from food consumption. Meat accounts for almost a third of the climate impacts from food consumption even if it accounts for less than ten percent of the total food flow. To substitute the protein demand now filled by bovine meat with chicken meat would reduce the total climate impact with 15 percent. Also some vegetables such as tomatoes and salads have a relatively large environmental impact. The calculated environmental impact from generated and treated food waste is presented in section 4. The climate impact of the avoidable food waste, i.e. food that could have been eaten if it had been treated differently, corresponds to 2 100 - 2 300 tons carbon dioxide equivalents. The climate impact from the treatment of food waste in the municipality of Lerum corresponded to -185 tons of carbon dioxide equivalents in 2011. In 2011 separately collected food waste in the municipality of Lerum was still composted and the mixed municipal waste incinerated. Currently the separately collected food waste in the municipality of Lerum is pre-treated and sent to anaerobic digestion. Given unchanged amounts of food waste, the climate impact from the current treatment of food waste would have been -280 tons of carbon dioxide equivalents. Assuming unchanged amounts of food waste and complete separate collection of food waste would result in -360 tons of carbon dioxide equivalents. These figures can be compared to almost 7 000 tons of carbon dioxide equivalents that the wasted food caused during production. This shows that prevention of food waste is a much more powerful measure in order to reduce environmental impact of food consumption than to change the treatment of the generated food waste. To combine material flow analysis from urban metabolism models with environmental performance from life cycle analyses is an interesting area of further research. The pilot study identified several knowledge gaps and areas to investigate further, e.g.: ? The amount as well as the quality of available LCA data needs to be considerably improved in order to be able to draw distinct conclusions from combinations of MFA and LCA data. ? It should be further investigated what level of aggregation is best suited for combining MFA and LCA data as well as whether it would make sense to use different levels of detail depending on different subareas. ? The applied MFA model, UMAn, does not fully account for private transports of goods across municipality borders. This results in certain uncertainties in the calculation of the environmental impacts from consumption in a specific municipality. The model should be adjusted to better reflect these private material flows across municipal borders. ? In cases where one material flow can be used for multiple purposes difficulties arise in terms of allocating the right subsets to the right purposes. It should be further investigated what material flows are relevant in this respect and how the subsets of these material flows should be allocated. ? Increased knowledge regarding the actual products in the waste streams is necessary in order to be able to accurately calculate the environmental impacts of waste treatment. For more complex products also knowledge regarding the content of the products is necessary. It is recommended that this is taken into account in future municipal waste picking analyses. ? The combination of MFA and LCA data stresses the importance that the chosen method in the MFA model is consistent, recognized and transparent since even small changes in material flows can get large effects regarding environmental impacts (e.g. for meat).