Interactions among environmental factors, management decisions, and field characteristics cause temporal and spatial variability in corn (Zea mays L.) and soybean [Glycine max (L.) Merr.] yields. The objectives of this paper are (i) to test whether yield response of corn to N and P and of soybean to P are spatially and temporally stable, and (ii) to evaluate the profitability of a variable rate (VR) N and P fertility management strategy over a 5-yr, corn-soybean rotation using this response information. A field near Windom, MN, USA, was cropped with corn (1997, 1999, and 2001) and soybean (1998, 2000). Three replications of 13 N and P treatments were established in a split-plot arrangement of a randomized complete block design. Treatments were applied at constant rates in strips across the entire field. Fertilizer N treatments were 0, 67, 112, 157, and 202 kg ha-1 and P treatments were 0, 56, and 112 kg P2O5 ha-1. The field was partitioned into sub-blocks for spatial analysis of yield response. Corn and soybean response to these inputs was estimated for each block, each year. Results indicate that spatial variation of crop response to these inputs is significant, and that response of corn and soybean to P is temporally stable in some parts of the field, but not others. Response to N was not temporally stable. Results of an ex post profitability analysis found that average returns over the 5-yr period from the VR N and P management strategy were $28 ha-1 higher than returns from a uniform application strategy.