We present a model of nucleon structure that is fully covariant. We begin with a Lagrangian that describes quarks coupled to various mesonic fields (\ensuremath{\sigma},\ensuremath{\pi},\ensuremath{\rho},\ensuremath{\omega}) and an additional field, \ensuremath{\chi}, which is required to bind the quarks into a nucleon. The couplings of the \ensuremath{\sigma}, \ensuremath{\pi}, \ensuremath{\rho}, and \ensuremath{\omega} fields to the quarks are fixed so as to reproduce the empirically determined coupling of these mesons to the nucleon. The latter couplings are taken from fits to nucleon-nucleon scattering data made using one-boson-exchange models. Therefore the free parameters of the model are the mass and coupling constant of the \ensuremath{\chi} field and the quark mass, ${m}_{q}$. In order to simplify the problem, the nucleon is assumed to virtually decay into a quark and a diquark. Equations which specify this amplitude are found by using the equations for the quark and meson field operators obtained from our Lagrangian. The equations which we solve are fully covariant and nonlinear and are solved by iteration. In this model the quark dynamics is governed by the mesonic fields whose source is the nucleon itself. The amplitudes for the emission of these fields by the nucleon depend upon the (nucleon)\ensuremath{\rightarrow}(quark + diquark) amplitudes whose structures we are attempting to determine. This model therefore requires a self-consistent solution and leads to the nonlinear equations noted above. At this point we have not calculated mesonic corrections to the nucleon observables such as the magnetic moments, form factors, and ${g}_{A}$, although we have included the effects of all the mesonic fields in the calculation of the nucleon mass. However, the calculation of nucleon observables, using current operators which contain only quark fields, yields a surprisingly good fit to electromagnetic form factors, magnetic moments, ${g}_{A}$, etc. (It is possible that some aspects of these results will be less satisfactory when mesonic corrections are calculated.) The model has the further virtue of generalizing the SU(6) quark model of the nucleon so as to be consistent with the covariance requirements of the theory of special relativity. It is clear that the major limitation of the model, other than the use of the diquark approximation, is the lack of a satisfactory description of the confinement mechanism as the \ensuremath{\chi} field serves to bind the system but does not actually confine the quarks.