The chlorophyll meter provides a simple, quick, and nondestructive method to estimate leaf N status of rice (Oryza sativa L.), but the linear relationship between leaf N concentration on a dry-weight basis (N(dw)) and the meter reading differs depending on developmental stage and genotype. The objective was to determine whether prediction of N(dw) with the chlorophyll meter can be improved by a simple correction for specific leaf weight (SLW). Leaf N status was estimated by a chlorophyll meter (SPAD-502) and measured directly by micro-Kjeldahl procedure. Specific leaf weight was calculated as the ratio of dry weight to leaf area. In one field study with 'IR72', measurements were taken at midtillering, panicle initiation, and flowering stages on the uppermost fully expanded leaves of both N-deficient and N-sufficient plants. There was a linear relationship between N(dw) and SPAD values at each stage, but regression lines differed significantly between growth stages. Based on pooled data from all stages, the degree of linear fit was poor (r2 = 0.49). Adjusting SPAD values for SLW (SPAD/SLW improved the prediction of N(dw) (r2 = 0.93). For another set of measurements made on the flag leaves of five genotypes grown in the field and greenhouse, prediction of N(dw) was also improved, from r2 = 0.51 based on SPAD values alone to r2 = 0.87 based on the SPAD/SLW ratio. These results demonstrate that SLW influences the prediction of N(dw) by the chlorophyll meter, and that the adjustment of SPAD values for SLW greatly increases the accuracy of the prediction. However, when SPAD values are adjusted for SLW, the chlorophyll meter's estimate of N(dw) is no longer as quick, simple, or nondestructive as the nonadjusted SPAD values.