Associations between nitrogen (N) management and losses with soil health indicators (SHI) are widely presumed but relatively untested. An on‐farm experiment conducted in central Illinois was conducted to test potential relationships of SHI with agroecosystem outcomes of maize (Zea mays L.) yield and N losses under bounding N fertilization rates of 168 and 252 kg/ha. Chemical (n = 19), physical (n = 11), and biological (n = 14) SHI were measured at 24 locations within a 30‐ha field at five timepoints (V3, V10, RT, R6, and post‐harvest). Yields did not necessarily reflect N‐fertilization rates, with lowest yields (14.5 Mg/ha) under 224 kg/ha. Flow‐weighted nitrate‐N concentration was significantly higher under 168 kg N/ha (10.6 mg/L) relative to higher application rates, though cumulative tile nitrate‐N loads were similar. SHI varied more by sampling location and time than by N application rate. Depending on the time of sampling, distinct SHI were related to yield and tile N losses. Total soil carbon and permanganate oxidizable carbon (POXC) best explained yield variation, whereas POXC and sand content best explained variation in nitrate‐N loss. Nematode indices helped explain variability in yield (Simpson, Shannon) and nitrate‐N losses (Maturity), supporting recent propositions to integrate nematode measures into soil health assessments. This study provides a basis for expanding to multiyear assessments of SHI linkages with nutrient losses and crop productivity in the North Central US.This article is protected by copyright. All rights reserved