Perovskite solar cells (PSCs) have attracted a substantial interest owing to a very fast achievement of efficiencies >20 % within only several years of development. However, for any solar cell to become technologically viable, two additional milestones have to be achieved alongside high efficiency: means of industrial production and good operational stability. The former being mostly the focus of the private sector and industrially-oriented research institutes, understanding degradation and improving the stability of PSCs has become one of the major research topics in the field of emerging photovoltaics. This is also the focus of this thesis, where I investigate different degradation mechanisms in PSCs with the aim of extending their long-term stability. This dissertation can be divided into four parts: in the first one, I show how, through a series of prototypes, I designed and built a dedicated setup to investigate stability of PSCs. I also show how I developed its variation to improve by 700 % the efficiency of measuring current-voltage characteristics of our solar cells. Subsequently, I describe the intrinsic instability of, at the time, state-of-the-art PSCs. I show how, along with my colleagues, we managed to achieve breakthrough room temperature stability coupled with record efficiency through incorporation of Cs into perovskite. Next, I describe how and why these solar cells suffer from irreversible degradation if left at elevated temperatures however. This is due to a vulnerability to Au diffusion from the electrode, through Spiro-MeOTAD hole transporting layer into the perovskite. I show how, by incorporation of Cr diffusion barrier I managed to circumvent this problem (albeit at considerable efficiency loss). Subsequently, I describe an alterna-tive solution to the problem by substituting Spiro-MeOTAD with PTAA - a polymeric hole transporting layer. This approach effectively stops Au diffusion without compromising device efficiency, which was at the same time improved by incorporating Rb into the perovskite. Finally, a third approach is presented: replacing the Au electrode with one based on carbon nanotubes. This gives away with using Au and PTAA - both prohibitively expensive materials - and hence paves the way towards facile and inexpensive fabrication of stable PSCs. In the third part, I describe the effects of mobile ions in the perovskite on the behaviour of PSCs. First, I show how they lead to a partial reversibility of losses in aged devices. This has potentially far-reaching consequences for the way stability measurements of PSCs are conducted and how their lifetime is reported. Next, I show how the ionic movement in the perovskite can lead to PSCs with certain architecture to work as high-gain photodetectors. This is enabled by piling up of ions at the interfaces within the devices, which modifies the energy levels within. Finally, in the last chapter I describe systematically, how different factors such as temperature, illumination, atmosphere, load on the device and cycling of the environmental conditions affect the stability of PSCs. Based on this, I discuss the important parameters to control when ageing PSCs, as the first attempt to bring the community to a consensus on how to measure stability of PSCs. This is urgently needed to streamline the efforts to create stable PSCs and to commercialize the technology.