An unpromoted ultrafine iron nano-particle catalyst was used for Fischer-Tropsch synthesis (FTS) in a CSTR at 270 degrees C, 175 psig, H-2/CO = 0.7, and a syngas space velocity of 3.0 sl/h/g Fe. Prior to FTS, the catalyst was activated in CO for 24 h which converted the initial hematite into a mixture of 85% chi-Fe5C2 and 15% magnetite, as found by Mossbauer measurement. The activated catalyst results in an initial high conversion (ca. 85%) of CO and H-2; however the conversions decreased to ca. 10% over about 400 h of synthesis time and after that remained nearly constant up to 600 h. Mossbauer and EELS measurement revealed that the catalyst deactivation was accompanied by gradual in situ re-oxidation of the catalyst from initial nearly pure chi-Fe5C2 phase to pure magnetite after 400 h of synthesis time. Experimental data indicates that the nucleation for carbide/oxide transformation may initiates at the center of the particle by water produced during FTS. Small amount of epsilon'-Fe2.2C phase was detected in some catalyst samples collected after 480 h of FTS which are believed to be generated by syngas during FTS. Particle size distribution (PSD) measurements indicate nano-scale growth of individual catalyst particle. Statistical average diameters were found to increase by a factor of 4 over 600 h of FTS. Large particles with the largest dimension larger than 150 nm were also observed. Chemical compositions of the larger particles were always found to be pure single crystal magnetite as revealed by EELS analysis. Small number of ultrafine carbide particles was identified in the catalyst samples collected during later period of FTS. The results suggest that carbide/oxide transformation and nano-scale growth of particles continues either in succession or at least simultaneously; but definitely not in the reverse order (in that case some larger carbide particles would have observed). EELS-STEM measurement reveals amorphous carbon rim of thickness 3-5 nm around some particles after activation and during FTS. Well ordered graphitic carbon layers on larger single crystal magnetite particles were found by EELS-STEM measurement. However the maximum thickness of the carbon (amorphous or graphitic) rim does not grow above 10 nm suggesting that the growths of particles are not due to carbon deposition.