We present airborne observations of fetch-limited waves generated during strong offshore winds, ranging between 10 and 25 m s⁻¹, in the Gulf of Tehuantepec. The measurements were collected over a wide range of fetches, from young to fully-developed seas, and include one- and two-dimensional wavenumber spectra, wind velocities, and atmospheric turbulent fluxes. The observed evolution of the wave field is compared to numerical simulations using a recently proposed wave dissipation formulation. In particular, we compare the observed and computed evolution of the directional spectra for fetch-limited conditions under realistic wind forcing. The thesis is composed of two main chapters, both of which are self contained and can be read independently of one another. The first chapter focuses on the analysis of the wind and wave measurements collected in the Gulf of Tehuantepec during February, 2004. It provides a description of the environmental conditions, the instrumentation, as well as the techniques used in the analysis for the characterization of the wind-wave spectrum with respect the wave age or fetch. The results are discussed in the context of previous observations and current numerical wind-wave prediction models. The second chapter describes the comparison of the measured wavenumber spectra with numerical simulations using state- of-the-art wind-wave models. It provides general background information on the physics and the development of numerical wind-wave models. The model is first implemented and tested under the idealized conditions of spectral evolution in one spatial direction with stationary uniform winds. The two-dimensional simulations are carried out over the Gulf of Tehuantepec for two research flights with the assumed steady wind forcing based on the airborne wind measurements, QuikSCAT scatterometer winds, and reanalysis model winds from NCEP/ NARR. The computed wave spectra are directly compared to the airborne observations over a wide range of wave ages. The comparison between the observations and the numerical simulations highlights some of the issues associated with the source term balance in the tail of the spectrum. The performance of the model suggests that the Alves and Banner (2003) dissipation is a good first step for realistic simulations of the wind-wave spectrum. The possible reasons for differences between the measurements and the simulations are discussed