Nitrogen (N), phosphorus (P), and potassium (K) fertilizer use in cotton production is important. Data from a long-term experiment initiated in 1972 were used to evaluate effects of N, P, and K fertilization on lint yield and lint quality of Upland cotton in Oklahoma. The experimental design was a randomized complete block with four replications. Eleven treatments containing different rates of N-P-K were evaluated for each of three cultivars (Paymaster 145, Paymaster HS26 and Paymaster 2326 BG/RR). Analysis of variance (ANOVA) was carried out initially using general linear models (GLM) procedure in SAS. Using this preliminary analysis, a quadratic plateau model for lint yield against N rates was evaluated for each cultivar using nonlinear (NLIN) procedure in SAS. Application of all three nutrients had some effect on lint yield, although most of the response was attributed to N (all cultivars) and to some extent P (Paymaster 2326 BG/RR and Paymaster HS26). The critical N rate for Paymaster 145, Paymaster HS26, and Paymaster 2326 BG/RR was 45, 45, and 67 kg N ha-1 with a corresponding plateau lint yield of 734, 1156, and 1468 kg ha-1, respectively. The results for fiber length indicate that K fertilization is the key to long fibers, while N rates greater than 90 kg ha-1 significantly reduce lint quality variables.