Experimental data is imperative to characterize the behavior of materials used in structural analysis models. The aim of this data paper is to establish a database of material-scale uniaxial experiments, primarily conducted at EPFL, to assist with simulating the monotonic and cyclic behavior of structural steel materials and components. The proposed database consists of stress-strain data from test specimens subjected to tensile straining or to one of nine cyclic load protocols consistent with earthquake loading. The database includes 353 experiments on test specimens fabricated from 11 steel materials from around the world and an iron-based shape memory alloy; the database will evolve as additional tests are conducted. A downsampling method is developed that ensures a difference in strain energy of less than 0.5% between the original and downsampled data while reducing the number of datapoints by 89 times, on average. Results for S355 steels, constituting 57% of the database, are quantitatively validated by comparing the average measured initial yield stress and elastic modulus with prior work. The initial yield stress for S355 steels is found to be between 305 and 430 MPa and the elastic modulus between 190 and 230 GPa; these results are consistent with prior values. Finally, examples illustrate how the database can be used to calibrate constitutive models used in nonlinear finite element analysis.