The efficacy of fluorescent daylight, green, and blue lamps in reducing bilirubin levels was compared in two groups of infants with nonhemolytic hyperbilirubinemia: healthy infants in a term nursery and infants in an intensive care unit. The decline of serum bilirubin concentration was most rapid with the blue lamps, with the duration of exposure required being significantly shorter than that with the other two types of light in both groups. The rate of decline over the first 24 hours, as well as the overall rate of decline for the whole duration of phototherapy, was also significantly greater with the blue lamps, the rate being about twice that for the green lamps in the infants in the intensive care unit; the daylight lamps were intermediate in efficacy. The daylight lamps permitted easy clinical monitoring with minimal side effects, whereas the green and blue lamps were equally disturbing to the attending personnel. The green lamps caused severe erythema and tanning in the initial 200 hours of phototherapy. It appears preferable to use either daylight lamps, which permit enhanced clinical monitoring with adequate efficacy, or special blue lamps, which provide maximal therapeutic effect, rather than green lamps, which offer neither.