A methodology based on the social participation through the use of resident diaries was applied to evaluate the odour annoyance in the surroundings of an industrial park in Belgium during one year. The studied area covers about 8 km2 and includes13 potential odour emitting facilities. A network of 44 residents was implied in the survey, from which 19 were particularly considered for a detailed analysis. The questionnaire aimed at providing twice-daily an odour strength rating on a 6-level scale together with an odour qualifier. The fact that the response rate corresponding to "no-odour" was high (79%) is particularly discussed. Some tests are proposed to check the plausibility of the responses, the consistency within clusters of residents and the individual performance of respondents to discriminate among odour ratings. The odour rose is presented as an attractive and visual tool, particularly suited in the case of multi-source areas, to map the different odour emissions, to point out the more worrying ones, some others creating less annoyance and possible new unpredicted ones. The resident diary method proves to be very interesting, especially in the case of an industrial area, provided that validation checks are performed to insure the data reliability.