The need for effective ultraviolet A (UVA) protection is increasing. Broad-spectrum sunscreens are being used to protect the skin against aging and in the treatment of photodermatoses, where UVA protection can be vital. They are also used by patients taking photosensitizing drugs, such as 8-methoxypsoralen, to protect their skin against solar UVA. This raises the question of what components in broad-spectrum sunscreens are the most necessary for optimal UVA protection. In the present study, the components that can be used are compared and evaluated using an animal method with the inhibition of skin edema induced by 8-methoxypsoralen plus UVA as the biological end point. The results of this method indicate that powders do not provide any significant UVA protection. This could be due to the use of 8-methoxypsoralen in the test so that particularly the shorter UVA range is evaluated. Powders reflect mainly the longer UVA wavelengths. The UVA protection as measured with this method seems to be similar for the benzophenone derivative and for the dibenzoylmethane derivative. The UVA protection factors obtained are compared with those obtained in human skin using phototoxic erythema and UVA-induced tanning as parameters.