This paper offers a summary of different crowdsourcing modes to collect geo-information on landscape perception & preferences and cultural ecosystem services. Crowdsourcing modes range from harvesting information passively transmitted by large groups on the web to actively engaging the crowd to generate data by using dedicated mobile apps and web-platforms. The latter, active crowdsourcing projects, were described in more detail by analysing the organizational variables of the twelve projects that were identified. Crowdsourcing has great potential to advance the field of landscape perception & preference research as it enables the in-situ collection of real-time, location-based data. One of the main limitations of reviewed active and passive crowdsourcing modes, lies in the fact that sample selection bias easily occurs and sample representativeness of any target population has been proven hard to achieve. Often crowdsourcing projects are implemented with a strong focus on technical aspects and content, but with insufficient attention for participant engagement. Projects would benefit from more inter- and transdisciplinary approaches and professionalizing campaigns, and thereby bringing participant engagement to the heart of the project. We recommend more attention to be placed towards awareness raising, diversification of formats and activities to reach a larger diversity of participants, structured tracking of performance indicators and learning from participants’ feedback. Such strategies aim at enhancing participation and reducing bias in participant selection, which constrains the usefulness of the results for research, planning and policy.