Predefined multitest chemistry panels (PMCPs) have constituted a large proportion of laboratory tests and patient charges, even in pediatric settings, despite the absence of documented clinical utility for PMCPs and the general availability of random access analyzers that do not require predefined test combinations. We eliminated PMCPs in our tertiary children's hospital but placed no other restrictions on ordering, and observed a 32.7% reduction in the number of automated chemistry tests ordered. All 23 tests in the previous PMCPs showed a decline in utilization, >50% for 8 of the tests and 20-50% for 13 others, and this change was sustained throughout an 8-month follow-up period. The total number of orders for one or more tests increased by 8.2%, but the variety of combinations that were ordered increased by 280%. The most substantial changes included a decrease in the number of orders for combinations of >15 tests, and increases in the number of orders for single tests and combinations of 2 to 5 tests. Orders for combinations identical to all of the former PMCPs declined, with the exception of the 4-test electrolyte panel. There was a marked decline in orders for a 7-test panel identical to the recently defined HCFA-AMA Basic Metabolic Panel, and orders for combinations identical to the HCFA-AMA Liver Function and Extended Metabolic panels were vanishingly rare and nonexistent, respectively. The calculated reduction in patient charges was much greater than actual cost savings, but the reduction in total tests and increase in the variety of test combinations suggest that significant savings can be realized if clinicians are encouraged to order only the tests or combinations they need without imposing procedural, financial, and regulatory burdens.