Installed photovoltaic nameplate power have been growing rapidly around the worldin the last few years. But how much energy is returned to society (i.e. net energy) by this technology, and which factors contribute the most to the amount of energy returned? The objective of this thesis was to examine the importance of certain inputs and outputs along the solar panel production chain and their effect on the energy return on (energy) investment (EROI) for crystalline wafer-based photovoltaics. A process-chain model was built using publicly available life-cycle inventory (LCI) datasets. This model has been kept simple in order to ensure transparency. Univariate sensitivity analysis for processes and multivariate case studies was then applied to the model. The results show that photovoltaic EROI values are very sensitive to assumptions regarding location and efficiency. The ability of solar panels to deliver net energy in northern regions of the earth is questionable. Solar cell wafer thickness have a large impact on EROI, with thinner wafers requiring less silicon material. Finding an alternative route for production of solar-grade silicon is also found to be of great importance, as is introduction of kerf loss recycling. Equal system sizes have been found to yield an primary EROI between approximately 5.5-19 depending on location and assumptions. This indicates that a generalized absolute EROI for photovoltaics may be of little use for decision-makers. Using the net energy cliff concept in relation to primary EROI found in this thesis shows that primary EROI rarely decreases to less than the threshold of 8:1 in univariate cases. Crystalline photovoltaics under similar system boundaries as those in the thesis model does not necessarily constrain economic growth on an energetic basis.