The purpose of this study was to compare two methods for maintaining intravenous therapy in preterm infants. A quasiexperimental research design was used. The model for the study was Roy's adaptation model (1). The population for this study was all infants, in neonatal intensive care units, undergoing intravenous therapy. A convenience sample of all infants admitted to one neonatal intensive care unit was used. Two hundred infants were the subjects. One group (n= 100) of infants received intravenous therapy via scalp vein needle, and the other group (n= 100) of infants received intravenous therapy via teflon catheter. A data collection form was attached to each subject's chart. The number of attempts necessary to insert the intravenous cannula, the duration of time in hours that the cannula was in situ, and the reasons for cannula removal were compared. Statistical analysis byt-tests and chi-squares showed that the infants in the teflon catheter group were younger gestationally, weighed less, and spent more time on the respirator, more days on intravenous therapy, and more days in the hospital (p<.0001). No significant difference was found in duration times between the two groups. Scalp vein needles primarily were removed for tissue edema, while teflon catheters were removed for tissue edema, cannula clotting, and cannula leakage (p<.007). These findings indicated that for this sample of preterm infants, the teflon catheter was the most efficient cannula for intravenous therapy.