Using microdata, that is, representative samples of 114,403 German long-term care dependants (LTCDs) observed from 2000 to 2009, we give a comprehensive insight into the length of stay (LOS) in long-term care (LTC). Furthermore, this paper evaluates the effects of longevity on the LOS, thus revisiting the debate on the validity of the competing theories of compression or expansion of morbidity in LTC. The analysis finds significant effects on the LOS when AGE is controlled for, albeit do not confirm the time-to-death hypothesis. However, controlling for ASSESSMENT LEVEL suggests an improved health status of LTCDs over time, thus supporting the time-to-death hypothesis. An analysis of the mortality rates of LTCDs is to give insight into the opposing results. But the regression of mortality shows a divergence in the development of mortality rates for different disability levels. This is evidence to suggest that the 'improved' health status in LTC is not only due to actual changes in the health status, but also a consequence of political intervention.