The fumigationextraction method to estimate microbial C was applied to a range of seven soils with organic-C contents up to 47%. The additional oxidizable organic-C released by CHCI, fumigation at -5 kPa water potential followed by extraction with 0.5 M KrSO,. was compared with the microbial C estimated by a modified substrate-induced-respiration (SIR) method. and by in situ labelling of cells with ["Clgiucose. The extractable organic-C flush comprised 3763% of the microbial C. The overall k,, factors. to convert from the organic-C flush to microbial C were 0.37 for the SIR method and 0.42 by "C-1abelling. The relationship between microbial C and the oxidizable C released by fumigation was similar for both organic and mineral soils. Accumulated data from 168 comparisons of the relationship between the organic-C flush and microbial C were m-examined. The It,,-factors varied widely between soils and were strongly influenced by soil water content. It is recommended that soils should be rewetted, when required, to the -5 kPa water potential. to ensure that the fumigation stage is fully effective. The reliability of the SIR and fumigation-incubation methods is discussed and a revised calibration for the SIR method: microbial C (!~g g-') = 50 (111 CO! g-' h-l) is proposed, based on "C-labelling and alternative methods of calculating the CO:-C flush when using the fumigation-incubation method. Taking account of possible underestimation of microbial C by these calibrations. a value of 0.35 is suggested as being representative of an overall &.-factor for New Zealand soils. Variability between soils means there is an error in applying an average factor to all soils, but a standard method should be adequate for most applications of microbial C measurements.