The theoretical mass specific capacitance (C-s) of polyaniline (PANI) is firstly estimated by combining electrical double-layer capacitance and pseudocapacitance. The maximum C-s is 2.0 x 10(3) Fg(-1) for one single PANI electrode. In present work, the PANI nanofiber modified stainless-steel (SS) electrode (PANI/SS) was used to assemble supercapacitors. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images indicate that the PANI nanofiber has a coarse surface arising from the heterogeneous structure which likes an aggregation of nanoparticles. The performance of the assembled PANI/SS supercapacitors was investigated by cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), and galvanostatic charge/discharge methods in 1.0 M H2SO4. The maximum C-s obtained from these methods in present work is 608, 445.0, and 524.9 Fg(-1), respectively, which is only 30%, 22%, and 26% of the theoretical one. The significant difference between the experimental and the theoretical value indicates that only a low percentage of PANI (effective) has contribution to capacitance. The percentage of effective PANI depends on both the diffusion of dopants (counter-anions) and the conductivity of PANI. Under practical conditions, the former factor makes PANI nanofiber behave like a concentric cable with only the shell part involved in the charge/discharge process. The latter one which determines the electron transfer rate in PANI has an influence on the degree of redox reaction. In present work, the heterogeneous structure of the PANI nanofiber has a negative effect on the conductivity. (C) 2009 Elsevier B.V. All rights reserved.