In this paper I present a new method for analyzing long series of observed sea levels. My method provides insight in sea level and the changes over time that are required by practicing engineers, assigned the design of new, modified or renovated structures along the coast. After extensively reviewing earlier research of sea level and tide, I present my method. The analysis relies on the application of classic harmonic analysis, which is made operational in a script in the programming language Python. Rigorous statistical testing is introduced to test the significance of trends and cycles in sea level and tide. This application of harmonic analysis and the introduction of formal statistical testing appear both to be new in this field. The method is tested on six locations in the Dutch North Sea, all with continuous sea level records of at least 130 years. The results of the statistical quality tests are shown in this paper. Subsequently, I show my findings related to mean sea level, lunar and solar tide and wind setup. Subsequently I analyze the long-term trends and cycles in Mean Sea Level. Long-period cycles with periods equal to the oceanic perigean and nodal tide are found to be important for a correct interpretation of sea level over time. Statistical tests show that acceleration of the rate of sea level is not significant up to 2021; the last year in the dataset. I compare my results with contemporary projections of sea level rise. The comparison reveals that in the Dutch North Sea the projected rates of rise are a factor two or more higher than the empirical rates established in this paper.