ABSTRACT: Based on measurements of phytoplankton biomass, absorption coefficients and photosynthetic performance obtained from samples collected in the subtropical Atlantic Ocean, standard sets of bio-optical parameters were selected and applied to aspectral model of primary production. Computed profiles of instantaneous production, integrated over the day, and spectral irradiance were compared with measurements of 12 h in situ production and in situ spectral irradiance to assess themodel's performance. Although the model performed well for the stations located on the western side of the transect, at the eastern side of the basin the model overestimated both primary production and irradiance below the mixed layer. The modelledirradiance profile was then reconciled with the observations of irradiance by increasing the assumed contribution to absorption by yellow substances. When the production model was re-implemented with the new irradiance values below the mixed layer, thecomputed and measured production profiles of all the stations were in good agreement. In the spectral model used in this study, the shape of the absorption spectrum of phytoplankton was used as a proxy for the shape of the photosynthetic action spectrum.Because non-photosynthetic pigments (NPPs) were abundant in pigment samples collected in the study region, the influence of NPPs on the shape of the absorption spectrum was examined and its effect on computations of primary production was quantified. Whentotal a_ph(z, λ) and photosynthetic a_ps(z, λ) phytoplankton absorption at wavelength λ and depth z were used toquantify the error resulting from failure to correct for the influence of NPPs on the shape of the absorption spectra, the results showed small errors in the computation of production at depth (up to 20%) and integrated, water-column primary production (amaximum of 10%).