Avisually reinforced operant paradigm was employed to examine the relationship between the difference limen (DL)for intensity and level of the standard during infancy.In Experiment 1, 7-month-old infants and adults detected increments in continuous noise presented via headphones at each of four levels ranging from 28 to 58 dB SPL. Noise stimuli were 2-octave bands centered at either 400 or 4000Hz, and increments were 10 and 100msec in duration. Infants' DLswere significantly larger than those of adult subjects and significantly larger for low-than for high-frequency stimuli. For the highfrequency noise band, infants' DLs were generally consistent with Weber's law, remaining essentially constant for standards higher than 28 dB SPL(3 dB SL)for 100-msecincrements and 38 dB SPL(13 dB SL)for lO-msecincrements. For low-frequency noise, infants' absolute thresholds were exceptionally high, and sensation levels ofthe standards were too low to adequately describe the relationship. In Experiment 2, 7-month-old infants detected 10-and 100-msec increments in 400-Hz noise stimuli presented in sound field. Infants' low-frequency DLswere large at low intensities and decreased with increases in level of the standard up to at least 30 dB SL. For both low-and high-frequency noise, the difference between DLsfor 10-and 100-msecincrements tended to be large at low levels of the standard and to decrease at higher levels. These results suggest that the relationship between the DLand levelof the standard varies with both stimulus frequencyand duration duringinfancy. However,stimulusdependent immaturities in increment detection may be most evident at levels within approximately 30 dB of absolute threshold.One ofthe most basic of auditory capacities is the ability to detect a change in the intensity or level ofan auditory stimulus. This ability has been widely studied in both normal and hearing-impaired adults, but little is known about discrimination of intensity differences in infants and young children. The few experiments examining this topic from a developmental perspective are consistent in showing poorer performance in young subjects. Sinnott and Aslin (1985) reported difference limens (DLs) ranging from 3 to 12 dB for a group of 7-to 9-month~old infants detecting increments in a constantly repeating 1-kHz tone presented at 60 dB SPL. By comparison, adult DLs obtained under identical conditions ranged from 1 to 2 dB. Bull, Eilers, and Oiler (1984) varied the intensity of individual synthetic syllables embedded within multi syllable wordlike units and found increment thresholds on the order of 2-6 dB in young infants; Schneider and Trehub (1985) reported similar estimates of2.5-5 dB for infants' detection of increments in continuous broadband noise. These values compare with DLs of less than 1 dB typically obtained in adult subjects for similar types of stimuli (see, e.g., Miller,