Municipalities wishing to develop collective composting i.e. at the bottom of collective buildings need methods and data to implement and evaluate such operations. These methods and data are for instance the production of organic waste (o.w.) in collective home per inhabitant and per year and the amounts of o.w. really diverted from the usual waste collection following the implementation of a collective composting operation. This paper aimed to propose such methods. These were tested at the territory of Rennes Métropole. Since there was no selective collection of o.w. at RM, measure organic waste production imposed to collect residual household waste (r.h.w.) then to quantify o.w. within. It also required to select the area of sampling, size the sample i.e. the number of apartments which production of o.w. should be collected, take into account the week frequency of usual collection then apply sampling at every season. The defined area corresponded to one part of urban area. It excluded city center in which economic activities concentrate in collective buildings. Taking into account the r.h.w. production at Rennes Metropole was 211 kg /inhbt /y in 2010 and that every waste sample should be at least equal to 2 tons, a sample of 610 apartments i.e. 1123 inhabitants was selected randomly from a list of 2500 addresses included in the defined area. The selected buildings were not equipped with collective composter. Since r.h.w. collection was practised twice a week on the selected area, i.e. on Monday and Thursday, two samples should be collected per week in parallel to usual collection. Four sampling campaigns were applied corresponding to winter, autumn, spring and summer. Then, the samples were characterized according to the French procedure MODECOM. The mean quantity of r.h.w. in collective homes was found around 145 kg/inhbt/y. and varying from 124 to 168 kg/inhbt/y. according to the season. Discard with global r.h.w. production i.e. 211 kg/inhbt/y should be investigated especially regards to its impact upon estimation of potential waste diversion through collective composting. The mean quantity of o.w. was found around 42.1 kg/inhbt/y. with a little variation over the seasons i.e. we found 10% of variation. The part of o.w. in the r.h.w. was found around 29% with a variation over the season of 10%. In spite of this low seasonal variation, the reduce of the number of sampling should be considered carefully. Three methods were tested in order to estimate o.w. diversion following the implementation of a composting operation. The first method (M1) was based on the weighing of all the o.w. from the building by a volunteer resident of the composting site. Thus, three collective composting sites were equipped with similar composters and monitored from 6 to 12 months. The second method (M2) consisted in the individual weighing of o.w.. Thirty households from different residences of Rennes Metropole weighed their o.w. at home, before the supply to the composter. A third method (M3) was based on the conservation of the mass of mineral matter during the composting process. Applying methods M1 and M2 led to respectively 43.5 and 47.4 kg/inhbt/y. Although, these are not significantly different from o.w. production (42.1 kg/inhbt/y.), it should be kept in mind that volunteers involved in the methods M1 and M2 might have higher production of o.w. due to their food habits. This assumption was supported by the results from the study carried out in the same framework at individual homes. Such assumption led to propose some rules allowing to estimate o.w. production thanks to whether M1 or M2 or on basis of sampling campaigns were proposed. M3 was found responsible for large errors in the estimation of waste diverted.