Ultraviolet (UV) radiation plays an important role in the ocean for the biogeochemical cycling through photooxidation of colored dissolved organic matter and metals. It also influences the release of trace gases into the atmosphere. Understanding interaction and climate feedback mechanisms of these processes requires global long-term monitoring of UV radiation in the ocean. Ocean color sensors provide the diffuse attenuation coefficient for downwelling solar irradiance (Kd) as an estimate for the mean light penetration depth in the surface ocean. However, Kd products for the UV spectral range are currently only indirectly inferred from measurements in the visible spectral region. We exploit the UV and visible wavelengths of the TROPOMI sensor onboard Sentinel-5P to directly infer diffuse attenuation coefficients in the UV and blue spectral range. Our approach is based on Differential Optical Absorption Spectroscopy in combination with radiative transfer modeling which has been used in earlier studies to successfully derive Kd in the blue from the vibrational Raman scattering (VRS) signal. VRS was detected in the backscattered radiances of TROPOMI-like sensors at a spectral resolution around 0.5 nm. We adapt this method for the TROPOMI sensor and more spectral regions to obtain two novel Kd products in the UV range (312.5–338.5 nm and 356.5–390 nm), additionally to the blue Kd (390–423 nm). One month of TROPOMI data show high sensitivity to retrieve VRS (fit errors <15%). The resulting Kd values in the UV and blue agree reasonably (RMSD <0.31 m−1) with in-situ data measured during a transatlantic cruise. Kd-blue compared to wavelength-converted Kd(490) products from OLCI and OC-CCI shows that the differences between the three data sets are within uncertainties given for the OC-CCI product. TROPOMI Kd-blue results have higher quality and much higher spatial coverage and resolution than previous ones from SCIAMACHY, GOME-2, and OMI. Perspectively, global TROPOMI Kd products, once thoroughly validated, can be used as input for modeling, e.g., photochemical reaction rates of climatically important compounds and inhibition of primary productivity.