Cotton pickers are now available to successfully harvest narrow-row cotton (Gossypium hirsutum L.) planted in many different row configurations. Research in Mississippi to determine optimum management techniques for the narrow-cotton production has been underway for several years. Factors investigated have included cultivars, nitrogen (N) rates, N application timing, growth regulators, planting dates, and planting patterns. A 6-year study (1989-1994) at the Delta Research and Extension Center located at Stoneville, Mississippi showed an average lint yield increase for ‘DES 119' of 5.3% (56 lb lint/A) with 30-in rows compared to 40-in rows when averaged across N management and growth regulator systems. The response ranged from a decrease of 46 lb lint/A (3.2%) to an increase of 112 lb lint/A (12.5%). Research with different planting patterns at the same time showed good potential for skip-row (2x1) planting with narrow-row systems. The first study with 30-in solid and skip-row planting patterns and increasing N rates showed no significant increase in lint yields above 90 lb N/A. Additional information was needed to better define the interactive role of N rates with respect to lint yield under dryland, rain-fed conditions. A 32-in row system (which allows equal picker efficiency, five 32-in rows vs four 40-in rows) with solid and skip-row (2x1) planting patterns were evaluated with N rates ranging from 60 to 165 lb N/cotton-acre (C-A) in 15-lb /A increments with all N applied prior to planting. The urea-ammonium nitrate (UAN) solution was "knifed" into both sides of the row with a ground-driven squeeze pump system that metered N to each row. For the skip-row areas, the UAN solution was by-passed from the knives back into the tank so that only the planted cotton rows were fertilized. A4-year study was initiated in 1996 on a Bosket very fine sandy loam (Mollic hapludalfs) with 16 treatments (2 planting patterns X 8 N rates) arranged in a randomized complete block design with five replications. Results from the first two years have been included in this report. Yields were reported in cotton-acre units even though these tended to over-emphasize differences when compared to land-acre yields. Lint yield per cotton-acre in 1996 ranged from a low of 665 to a high of 805 lb/C-A for solid planted cotton and 849 to 1052 lb/C-A for the 2x1 skip-row pattern. There were no significant differences with respect to N rate for the solid planted cotton. In the skip-row system, the highest yield was obtained with 120 lb N/C-A. However, lint yields were quite varied among N rates. The lint response to skip-row planting in 1996 ranged from 15 to 58% depending on the N rate with no specific pattern identified. For 1997, lint yield ranged from 977 to 1141 lb/C-A for solid planted 32-in row cotton and 1248 to 1451 lb/C-A for skip-row cotton. The skip-row response ranged from 18 to 36% and again no significant response to N rates. When averaged across the two years, there was no response to increasing N rates and a skip-row response of 167 to 348 lb lint/C-A (19 to 38%). When averaged across all N rates, the skip-row response was 214 lb lint/C-A (28.9%) and 282 lb lint/C-A (26.6%) for 1996 and 1997, respectively. Increasing N rates did delay maturity as determined by the percent first harvest but the differences were small. In summary, the data suggest that, in this non-irrigated environment, N is not a limiting factor with respect to lint yield. In going to a skip-row planting pattern, lint yields were increased by 19 to 38% while utilizing the same N rate per planted cotton acre. Further research is needed to evaluate alternate planting patterns to maintain the same number of row-feet of cotton per acre compared to solid plantings while increasing light in the lower canopy.