This paper reports the preliminary results of a two year study of the cotton simulation model GOSSYM-COMAX. Multiple plant mappings (fruit location records collected throughout the season) can be used to adjust simulations by GOSSYM. The objectives of this study were to establish the optimal number and the timing of plant maps to obtain the most accurate simulation by GOSSYM for cotton grown under varying field conditions. The data presented are from two locations observed in the Cotton Belt. DPL 50 was grown in two diverse environments, the Mississippi Delta (MS), its native environment, and Artesia, NM (NM), an extrinsic environment. Actual crop data (plant height, number of nodes, squares, bolls and yield) is compared to data simulated by GOSSYM to calculate an "agreement index". By examining the affect of different mapping treatments on simulations of the same variety in two diverse environments it is possible to establish how well GOSSYM is able to simulate a variety grown in the native environment vs. an extrinsic environment. Agreement levels were much lower and more variable in comparisons made with NM data than with MS data. This suggests GOSSYM does not simulate a variety out of its 'natural' environment as well as one in its native environment. Regardless of location, mapping treatments which include post bloom mapping gave the highest levels of agreement.