The Pre-Aerosol, Cloud, and ocean Ecosystem (PACE) mission will carry into space a spectrometer measuring at 5 nm resolution in the ultraviolet (UV) to near infrared (NIR) and at lower resolution in spectral bands in the NIR and shortwave infrared (SWIR). These observations have great potential for improving estimates of marine reflectance in the post-EOS era. In view of this, we evaluate, using simulations with a coupled radiation transfer code, the gain in marine reflectance accuracy expected by including observations in the UV and SWIR compared with just using observations in the visible to NIR. The study is performed for the PACE threshold aggregate bands with respect to the standard set of bands used to generate ocean color products. The top-of-atmosphere (TOA) signal measured by the PACE spectrometer is simulated for a variety of realistic atmospheric and oceanic conditions. The TOA reflectance and the marine reflectance of the simulated ensemble are decomposed into principal components, and the components of the TOA reflectance sensitive to the ocean signal identified. Inverse models are constructed to retrieve the principal components of the marine reflectance, allowing a reconstruction, therefore an estimation of the marine reflectance. Theoretical performance is quantified as a function of angular geometry, aerosol properties, and water type, showing a significant improvement in retrieval accuracy when using the extended spectral range. On average over all the situations considered (including sun glint), the RMS error is reduced from 0.0037 to 0.0024 at 412 nm, from 0.0013 to 0.0007 at 665 nm, and from 0.0010 to 0.0004 at 865 nm (Case 2 waters are better handled). The performance is degraded at large zenith angles and aerosol optical thickness, is better at scattering angles around 120-130 degrees, and exhibits little dependence on aerosol single scattering albedo and aerosol scale height.