Ultraviolet-B (UV-B; 280–320 nm)–emitting lamps unavoidably emit ultraviolet-A (UV-A; 320–400 nm) and ultraviolet-C (UV-C; <280 nm) radiation. Short-wavelength–blocking filters are generally used to limit the wave bands of UV under investigation. The widespread use of such filters means that all exposures to UV-B radiation will have a significant UV-A component. Therefore, the physiological effects unique to UV-B exposure are difficult to clearly isolate. This study presents a method to remove the UV-A and UV-C “contamination” using a liquid potassium chromate (K₂CrO₄) filter, thus allowing more direct assessment of the effects of UV-B exposure. Cultures of the green marine alga Dunaliella tertiolecta were grown in the absence of UV radiation. Sunlamps supplied the UV radiation for a 24 h exposure (solar radiation was not used in this study). The UV radiation was filtered either by the standard method (i.e. cellulose acetate (CA) with polyester = Mylar controls) or by a liquid filter of potassium chromate. Photosynthetic responses were compared. Major decreases in the ratio of variable to maximal fluorescence in dark-adapted cells and photosynthetic capacity were observed in CA-filtered cultures, whereas no change was observed in cells exposed to the same UV-B flux with the UV-A removed by K₂CrO₄. The use of a CA filter with a Mylar control does not link results unequivocally to UV-B radiation. Such results should be interpreted with caution.