Introduction In constructing new dwellings or restructuring existing ones, needs and preferences of households for which these houses are intended, have to be taken into account. Preferences can be measured using conjoint measurement. The goal of this study is to explore whether preferences for dwellings are dependent upon the way in which these dwellings are presented in a conjoint measurement, i.e., using ¿text only¿, using ¿¿text and color photo¿ or using ¿text and black-and-white impression¿. Methods and respondents We developed eight dwelling profiles based on combinations of characteristics of the dwelling and the residential environment. We presented these profiles in three different ways to 28 respondents. Each respondent was asked to rate the 24 profiles and to indicate whether they would want to move to the particular dwelling. After the task, respondents filled out a short questionnaire and were asked to participate in a semi-structured interview about the consistency of their choices. Results For five of the eight profiles, the ratings differed between the three presentation methods. Significant differences in choices were observed for one of the eight profiles only. The inclusion of images in the conjoint analyses led to a number of differences between the models. These findings can be explained by elements on the images that are not systematically varied and thus have a disturbing influence. This conclusion was supported by the post-test interviews and questionnaires; many respondents indicated that they had been influenced by details on the images, such as the color of the brick. Conclusion Respondents seem to be influenced by details on photos and impressions. If images have to be included in a conjoint measurement task, it is of utmost importance to make sure that all potentially disturbing details are cleared away from the photos or impressions.@en