The purpose of this study was to compare the amount of total milk intake, feeding time, sucking efficiency, heart rate (HR), respiratory rate (RR), and oxygen saturation (SpO2) of premature infants when fed with either signal-hole or cross-cut nipple units. Twenty stable infants admitted to a level II nursery in a tertiary care center with gestational ages averaging 32.2+/-3.2 wks were enrolled. Subjects had an average postmenstrual age of 34.1+/-1.6 wks, and average body weight of 1996+/-112 gm. A crossover design was used and infants were observed for two consecutive meals separated by a four-hour interval. They were bottle fed with equal feeding amounts using a single-hole and cross-cut nipple administered in random order. Results showed that infants fed with single-hole nipple units took more milk (57.5+/-8.3 ml vs. 51.6+/-9.5 ml, p=.011), had a shorter feeding time per meal (11.5+/-4.9 min vs. 20.9+/-5.0 min, p<.001), and sucked more efficiently (5.8+/-2.5 ml/min vs. 2.7+/-1.0 ml/min, p<.001) compared to those fed through cross-cut nipples. Infants using cross-cut nipple units had a higher RR (44.4+/-4.6 breaths/minutes vs. 40.8+/-4.9 breaths/minutes, p=.002) and SpO2 (96.1+/-1.4% vs. 94.6+/-3.2%, p=.044) than those using single-hole nipples. Oxygen desaturation (SpO2<90% and lasting for longer than 20 sec) and bradycardia were not recorded in either group of infants during feeding. Compared to using cross-cut nipple units, premature infants using single-hole nipple units take more milk and tend to tolerate feedings better. A single-hole nipple may be a choice for physiologically stable bottle-fed premature infants.