Polyvinyl alcohol (PVA) as a gel matrix was used to coimmobilize nitrifying and denitrifying bacteria by repeated freezing and thawing. Factors affecting single-stage biodenitrification (simultaneously occurring nitrification and denitrification in a reactor) process such as nitrifying bacteria and denitrifying bacteria ratio, organic carbon source, pH, alkalinity, temperature, dissolved oxygen (DO), and the operational stability of coimmobilized cells were investigated. Besides, the kinetics model of single-stage biodenitrification was studied. The experimental results showed that nitrogen removal rate was most fast when nitrifying to denitrifying bacteria ratio was between 1.5:1 and 3.6: 1; in four carbon sources of methanol, ethanol, acetic acid and glucose, nitrogen removal rate was most fast when using ethanol; the optimal values of temperature, pH and DO were 30degreesC, 8.2 and 2similar to6 mg(.)l(-1), respectively; the higher alkalinity to ammonia ratio was, the faster nitrogen removal rate was, while nitrogen removal rate went steady when alkalinity to ammonia ratio is over 9. The continuous operational system remained stable for a period of more then 60 days, and the coimmobilized cells could easily adapt to an increase in loading. The effect of internal diffusion on overall rate could be neglected, and the limiting rate process was biochemical reaction in the single-stage biodenitrification process using coimmobilized cells. The reaction kinetics of coimmobilized cells followed Michaelis-Menten form, and its kinetics constant, K-m and v(max) were 303.0 mg(.)l(-1) and 0.096 mg(.)l(-1) (immobilized cells)(.)s(-1), respectively. The relative errors of predicted and experimental nitrogen removal rate were almost all less than 10%.