We studied the short-term and long-term effects of transient asymptomatic neonatal hyperammonemia on neurologic function in 21 preterm infants with normal ammonium levels and 25 with hyperammonemia (range 40 to 72 mumol/L) during the first weeks of life. The hyperammonemic infants were prospectively randomized to treatment with orally administered arginine free base 1 to 2 mmol/kg/day for 2 months (n = 13) or to a no-treatment control group (n = 12). Cortical function was assessed by auditory response and habituation during the first month of life. An auditory response was shown by 64% of the hyperammonemic infants and 43% of the normoammonemic infants (P not significant). Plasma ammonium levels at the time of examination bore no consistent relationship to whether an infant responded to an auditory stimulus. Number of trials to reach auditory habituation was also not different, and plasma ammonium level did not correlate with the presence or absence of habituation. IQ testing at 6, 12, 18, and 30 months showed no significant differences between groups. Early plasma ammonium levels did not have an effect on 30-month IQ scores. These findings suggest that transient asymptomatic hyperammonemia in premature infants is not associated with short-term or long-term neurologic deficits through 30 months of age. This study does not support the need for treatment of transient asymptomatic hyperammonemia in the premature infant.