Many fish species require to move along rivers to complete their life cycles. Therefore, they are one of the animal groups most affected by the intensive use that humans make from rivers. Among all the impacts, the installation of transversal obstacles to the river is one of the most notable alterations affecting fish movements. The best solution to recover the free movement of fish is to eliminate the obstacles. However, their social benefits make inviable their removal and often, the only way to restore the longitudinal connectivity is by building fish passes or fishways. There are many types of fish passes, nevertheless, due to their versatility and ability to deal with a wide range of different types of obstacles, stepped fishways are the most common alternative. Despite their attractiveness to enable the free movement of fish, stepped fishways are sensitive to the natural variability of rivers and their performance can be easily altered by the variable boundary conditions of rivers. This thesis is a systematic study of the effects of variable boundary conditions in stepped fishways. To do this, the hydraulics of different types and subtypes of fishways are studied, taking into account the most extended calculation methodologies, defining their limitations and proposing new calculation methods. Field study cases under different boundary conditions together with cases from specialized references are considered, resulting in a general methodology for the mean water level modelling of stepped fishway. The developed methodology allows to consider the natural variability of rivers in fishway design projects which will ensure the correct performance during the whole hydrological period and the performance optimization of those fishways already constructed. Likewise, it allows to consider variability of boundary conditions in fishway assessments, which will produce more relevant result and conclusions in those studies carried out over a long-time period subject hydrological variability. List of original articles 8 List of original articles This thesis is based on three original works, which constitute the main body of the thesis. All of them have been published in different indexed international journals. The author of the thesis has been the first author of the three articles, responsible of proposing the initial idea, planning and conducting experiments, designing the methodology, analyzing the data, discussing the results and writing the papers. Author order, title of the publication and journal information are presented below: