This paper is based on the keynote lecture given at WILA 10 in Halden. It focuses on data collection among Norwegian-American heritage speakers and how such data is made available through an online database. The first part of the paper gives a brief overview of data collection documenting Norwegian as a heritage language in America from the early 1930s and towards the end of the millennium. The second part describes the data collection done during the last decade, when the author organized a number of field trips in the Norwegian-American communities. The last part of the paper demonstrates some of the possibilities that CANS (Corpus of American Nordic Speech) makes available for the research community.