We hypothesized that when workload is heavy in the neonatal intensive care unit (NICU) housestaff caring for the neonates may have less time and therefore be less likely to perform certain procedures, specifically lumbar punctures (LPs) and phlebotomy. Although initially time consuming, the procedure of umbilical artery catheterization (UACs) will eventually save time(i.e. less manual blood draws), and therefore, may be more likely to be performed by housestaff when workload is heavy in the NICU. Intubation is usually not elective, and we hypothesized this procedure would not be affected by housestaff workload. We designed a prospective cohort study over one academic year (1993-94), observing 785 infants exposed to 31 different housestaff working under various levels of workload. Workload was operationalized as number of NICU infants cared for by the houseofficer each night on-call. The procedures of interest were number of UACs, intubations, LPs, and peripheral phlebotomy performed by the houseofficer on-call. Using multiple linear regression, controlling for daily severity-of-illness of each NICU infant, the experience of the houseofficer on-call and the individual on-call attending, we found increased workload (number of NICU infants) resulted in a significantly greater probability that admitted infants received an umbilical artery catheter (p=0.02), and less probability that NICU infants received a lumbar puncture (p=0.0001) or peripheral phlebotomy (p=0.0002). The decision to intubate an infant was not affected by workload. Examining different levels of workload, infants admitted when there was less than 6 other NICU infants had a 21% probability of receiving a UAC, but a 38% probability when there was greater than 10 other NICU infants. Infants in the NICU when there was less than 6 other infants in the NICU had a 61% chance of being phlebotomized and an 18% chance of receiving an LP, but only a 20% chance of phlebotomy and a 4% chance of LP when there were greater than 10 other NICU infants. Thus, for equivalently severely-ill infants, with high workload there is a greater chance of receiving a UAC and less chance of being phlebotomized or receiving an LP. Therefore, we conclude workload in the NICU can affect decisions to perform procedures on infants.