Network Analysis for Regional Information (NARI) and the Cost-Effectiveness Procedure were tested by applying them to stream-gaging networks in Idaho. NARI was used to determine network design strategies that would maximize the value of additional data. Value of data was measured as the decrease in the probable true standards error of regional regression equations. NARI indicated that no significant decrease in regression error can be achieved by the collection of additional data and that better models should be sought. No major modifications to NARI are necessary to make it widely applicable. The Cost-Effectiveness Procedure was used to determine optimal network operation strategies. It showed network uncertainty can be reduced when six- or one-visit per year minimum constraints are in force. Sensitivity to various cost factors was examined. Attempts to model networks that included sites for collection of groundwater and water-quality data were unsuccessful. (USGS)