A spectroscopic diagnostic tool has been developed to determine the electron temperature and the neutral density in helium, hydrogen and argon plasmas from absolutely calibrated spectroscopic measurements. For each gas, a method of analysis which uses models specific to each species present in the plasma (neutral atom or singly ionized atom) has been defined. The experimental electron density is used as an input parameter to the models, and the absolutely calibrated spectroscopic data are processed beforehand to obtain the populations of the upper excited levels corresponding to the observed spectral lines. For helium plasmas, the electron temperature is inferred from the experimental helium ion excited level p = 4 population using a corona model, and then the neutral density is determined from the experimental helium neutral excited level populations using a collisional-radiative model for helium neutrals. For hydrogen plasmas, combinations of the electron temperature and the neutral density are determined from the experimental hydrogen neutral excited level populations using a collisional-radiative model specific to hydrogen atoms. For argon plasmas, the electron temperature is inferred from the experimental argon ion excited level populations using a collisional-radiative model for argon ions, and then the neutral density is determined from the experimental argon neutral excited level populations using a collisional-radiative model for argon neutrals. This diagnostic tool was applied to three experiments with different geometries and plasma conditions to test the validity of each data analysis method. The helium and hydrogen data analysis methods were tested and validated on helium and hydrogen plasmas produced in the VASIMR experiment, a plasma propulsion system concept. They gave electron temperatures and neutral densities that were consistent with other diagnostics and theory. The argon diagnostic tool was tested on argon plasmas produced in the VASIMR experiment, the Helimak experiment and the Helicon experiment. The electron temperature and neutral density obtained on both the Helimak and the Helicon experiments were consistent with other diagnostics and with theory, and validated the method of analysis. An impurity problem on the VASIMR experiment made it difficult for the data analysis to be validated.