A model to describe the evolution of conductivity measurements in emulsion polymerizations was presented in the first part of this work [Santos et al., J Appl Polym Sci, 2002]. The model was built on the assumption that surfactant is partitioned among the three principle phases of the polymerizing latex. Model parameters were estimated by fitting the model responses to experimental conductivity data obtained in different runs performed with similar solids content, reactor temperature, and initiator concentration. A soft-sensor strategy was then proposed for monitoring of the number of polymer particles (N-P) by combining the conductivity model with the available conversion, temperature, and conductivity signals. The main objective of the current work is to validate the conductivity model in a broader range of operation conditions, to follow the evolution of N-P (nucleation/ coagulation) on-line under different reaction conditions. A series of batch and semi-batch polymerization runs were carried out aiming to evaluate the model performance when different solids content, surfactant concentration, and reactor temperatures are used. Additionally, simulation studies were performed to assess the model sensitivity to the fluctuations of the conversion and temperature signals. Results showed that the model is able to perform accurate predictions of N-P even when disturbances of +/-2degreesC in temperature and +/-0.03 in monomer conversion take place during the polymerization process. (C) 2003 Wiley Periodicals, Inc. J Appl Polym Sci 91: 941-952, 2004.