In continuation of our recent efforts to model the acute toxicity of 56 phenylsulfonyl carboxylates to Vibrio fischeri using principal component factor analysis, the present paper deals with modeling of the same data set with extended topochemical atom (ETA) indices using genetic function approximation (GFA) as the statistical tool. The statistical quality of the best model using ETA descriptors is as follows: n=56, Q2=0.771, Ra2=0.861, R=0.935, F=69.0 (df5,50), s=0.172, AVRES=0.128. This equation was better in statistical quality than that obtained previously using principal component factor analysis as the data-preprocessing step. An attempt was also made to model the data set with different nonETA parameters (topological indices including Wiener, Hosoya Z, molecular connectivity, kappa shape, Balaban J and E-State parameters apart from physicochemical parameters like AlogP98, MolRef and H_bond_acceptor) and the best model showed the following quality: n=56, Q2=0.805, Ra2=0.820, R=0.913, F=53.5 (df4,51), s=0.196, AVRES=0.136. An attempt to use both ETA and nonETA parameters lead to an equation which is marginally better than the best ETA model: n=56, Q2=0.779, Ra2=0.865, R=0.937, F=71.7 (df5,50), s=0.169, AVRES=0.127. Use of the ETA indices suggested negative contributions of steric bulk, functionality of C10, volume of substituents on C10 and functionalities of chloro and nitro substituents at X1 position and positive contributions of functionalities of C4 and C1 and presence of substituents with electronegative atoms, especially at R2 and R3 positions. Furthermore, absence of substituents at R2 and R3 positions decreases toxicity. This study corroborates the results obtained from principal component factor analysis and suggests that ETA parameters are sufficiently rich in chemical information to encode the structural features contributing significantly to the acute toxicity of phenylsulfonyl carboxylates to V. fischeri.