Measurements of biologically active UV radiation made by the National Science Foundation (NSF) scanning spectroradiometer (UV-monitor) at Palmer Station, Antarctica, during the Austral springs of 1988, 1989, and 1990 are presented and compared. Column ozone abundance above Palmer Station is computed from these measurements using a multiple wavelength algorithm. Two contrasting action spectra (biological weighting functions) are used to estimate the biologically relevant dose from the spectral measurements: a standard weighting function for damage to DNA, and a new action spectrum representing the potential for photosynthesis inhibition in Antarctic phytoplankton. The former weights only UV-B wavelengths (280--320 nm) and gives the most weight to wavelengths shorter than 300 nm, while the latter includes large contributions out to 355 nm. The latter is the result of recent Antarctic field work and is relevant in that phytoplankton constitute the base of the Antarctic food web. The modest ozone hole of 1988, in which the ozone abundance above Palmer Station never fell below 200 Dobson units (DU), brought about summerlike doses of DNA-effective UV radiation 2 months early, but UV doses which could inhibit photosynthesis in phytoplankton did not exceed a clear-sky ''maximum normal'' dose for that time of year. The severe ozone holes of 1989 and 1990, in which the ozone abundance regularly fell below 200 DU, brought about increases in UV surface irradiance weighted by either action spectrum. Ozone abundances and dose-weighted irradiances provided by the NSF UV-monitor are used to derive the radiation amplification factors (RAFs) for both DNA-effective irradiance and phytoplankton-effective irradiance. The RAF for DNA-effective irradiance is nonlinear in ozone abundance and is in excess of the popular ''two for one'' rule, while the RAF for phytoplankton-effective irradiance approximately follows a ''one for one'' rule. ¿ American Geophysical Union 1992