In 1976 we reported a change in spectral distribution of "blacklight" fluorescent lamps. It was not possible to determine the spectral composition of these lamps by any codes or packaging materials. Phototherapy booths utilizing standard BL-HO lamps will accept lamps of at least two spectral distributions; both types are commercially available. This study was conducted to determine the biologic efficacy of these two lamp types that we refer to as BL-O and BL-N. The BL-O spectrum had a peak emission at approximately 350 nm with 98% of the energy between 320 and 400 nm. The BL-N spectrum had its peak emission at 365 nm with a range from 340 to 400 nm. The BL-O spectrum was at least 2.5 to 4 times as effective as BL-N in causing minimally perceptible phototoxicity in albino hairless mice given oral doses of 8 mg/kg of 8-methoxypsoralen. Food and Drug Administration (FDA)-approved specifications imply that the BL-O spectrum is to be used for psoralens and ultraviolet A (PUVA) phototherapy. If lamps with the BL-N spectrum are replaced by lamps with the BL-O spectrum, the metered dose must be reduced to no more than one-fourth of the previous dose or the patient may suffer serious phototoxic reactions.