This experiment evaluated alternative methods of soil preparation for cotton (Gossypium hirsutum L.) on a clay soil prone to compaction from traffic or tillage of wet soil. The treatments compared were: (1) a cotton-wheat rotation, where soil was maintained in good physical condition; (2) continuous cotton with soil preparation by minimum tillage (min-till); (3) continuous cotton with complete soil disturbance (max-till). In addition, as nitrogen (N) fertilizer application rates and application strategies may change with the method of soil preparation, we evaluated different methods of N application (anhydrous ammonia or urea applied before or after sowing, and placed under the crop row or between rows) with rates from 0 to 225 kg N ha-1. The experiment was repeated in 3 consecutive years. Max-till decreased soil nitrate, increased soil bulk density, reduced root development and, in some years, reduced water extraction in comparison with min-till and rotation cotton. Overall, soil under the max-till treatment was the most compacted; this was the result of soil preparation under wet conditions in the last two seasons. Cotton grown in rotation with wheat gave a yield that was, on average, 1.3% greater than that for min-till and 7. 1% greater than for max-till. The difference between rotation and min-till was small in the third year of the experiment, indicating that min-till may take time to show production advantages. Optimum N fertilizer rates averaged 145 kg ha-1 for rotation, 189 kg ha-1 for min-till and 210 kg ha-1 for max-till. Crop maturity was delayed slightly and micronaire reduced slightly in rotation cotton. Continuous cotton had a lower level of nitrogen uptake than did cotton grown in rotation with wheat; this is consistent with soil nitrate tests taken before sowing. The best fertilizer application strategy was to apply anhydrous ammonia under the crop row 2-4 months before sowing. The response to urea was similar, but side-dressing all N fertilizer after sowing was the worst method of application, particularly under max-till. We conclude that min-till is a viable option for soil preparation on this soil type, particularly under wet conditions. Weed, insect and disease factors may affect the decision to change rotation and tillage practice.