Using data from the German Socio-Economic Panel Survey (SOEP), this paper assesses the relationship between life satisfaction and religious practice. The main new result here is longitudinal. It is shown that individuals who become more religious over time record long term gains in life satisfaction, while those who become less religious record long term losses. This result holds net of the effects of personality traits, and also in fixed effects panel models. The paper has significant implications for the dominant, paradigm theory in SWB research, namely set-point theory. This theory holds that the long term SWB of adult individuals is stable, because SWB depends on personality traits and other stable genetic factors. It is already clear from the German panel data that about 20% of the population have recorded large long term changes in SWB. New evidence in this paper and elsewhere about the effects of consciously chosen life goals, including religious ones, on SWB is hard to reconcile with set-point theory. It is more in line with authentic happiness theory.