Our data show that the trends of methane have decelerated rapidly during the last decade. We argue that cattle, domestic animals and rice, that caused major increases of methane over the last century, are not major contributors to methane increases during the last decade and are not likely to contribute much to the increases of methane in the future. If this is so, the sources of past increases, related to agriculture and closely tied to increase of population, are getting uncoupled from the present and future increases of methane making it impossible to predict future concentrations based on past changes. New sources that are becoming important are not closely tied to population growth and have an unpredictable future. The calculations show that little or no increases of annual emissions are required during the last der-ade to explain the observed concentrations and trends. The rapidly declining trends have taught us at least one lesson - that we cannot project measured atmospheric trends into the future with any degree of credibility.