We previously reported on the mineralization of 2,4-dinitrotoluene (2,4-DNT) and 2,6-dinitrotoluene (2,6-DNT) in an aerobic fluidized-bed bioreactor (FBBR) (Lendenmann et al. 1998 Environ Sci Technol 32:82-87). The current study examines the kinetics of 2,4-DNT and 2,6-DNT mineralization at increasing loading rates in the FBBR with the goal of obtaining system-independent kinetic parameters. At each steady state, the FBBR was subjected to a set of transient load experiments in which substrate flux in the biofilm and bulk substrate concentrations were measured. The pseudo-steady-state data were used to estimate the biokinetic parameters for 2,4-DNT and 2,6-DNT removal using a mechanistic mathematical biofilm model and a routine that minimized the sum of the squared residuals (RSS). Estimated kinetic parameters varied slightly for each steady-state; retrieved parameters for q(m) were 0.83 to 0.98 g DNT/g XCOD d for 2,4-DNT removal and 0.14 to 0.33 g DNT/g XCOD d for 2,6-DNT removal. K-s values for 2,4-DNT removal (0.029 to 0.36 g DNT/m(3)) were consistently lower than K-s values for 2,6-DNT removal (0.21 to 0.84 g DNT/m(3)). A new approach was introduced to estimate the fundamental biofilm kinetic parameter S-b,S-min* from steady-state performance information. Values of S-b,S-min* indicated that the FBBR performance was limited by growth potential. Adequate performance of the examined FBBR technology at higher loading rates will depend on an improvement in the growth potential. The obtained kinetic parameters, q(m), K-s, and S-b,S-min*, can be used to aid in the design of aerobic FBBRs treating waters containing DNT mixtures.