A new methodology for soluble oxalic acid determination in grass samples was developed using a two enzyme reactor in an FIA system. The reactor consisted of 3 U of oxalate oxidase and 100 U of peroxidase immobilized on Sorghum vulgare seeds activated with glutaraldehyde. The carbon dioxide was monitored spectrophotometrically, after reacting with an acid-base indicator (Bromocresol Purple) after it permeated through a PTFE membrane. A linear response range was observed between 0.25 and 1.00 mmol l(-1) of oxalic acid; the data was fit by the equation A = -0.8(+/-1.5)+57.2(+/-2.5)[oxalate], with a correlation coefficient of 0.9971 and a relative standard deviation of 2% for n=5. The variance for a 0.25 mmol l(-1) oxalic acid standard solution was lower than 4% for 11 measurements. The FIA system allows analysis of 20 samples per hour without prior treatment. The proposed method showed a good correlation with that of the Sigma Kit.123443447