Development of C budget models for forests requires primarily, data on C stocks and fluxes in biomass, detritus, and soil organic matter. A secondary requirement is validation of model pool sizes and turnover times, ideally by isolating and characterizing the particular soil fractions corresponding to these pools. General development of process models is also enhanced by increasing our understanding of mechanisms that control the competing processes of decomposition and stabilization of soil organic matter. Forests of coastal Vancouver Island, British Columbia, have high soil and detrital C stocks, and prior to the onset of logging early in the 20th century were naturally disturbed, mainly by fire on the east side, and windthrow on the wetter, slightly cooler west side. The Coastal Forest Chronosequence (CFC) project comprises a broad range of studies on carbon stocks and fluxes, ecosystem function, and biodiversity, with eight chronosequence sites established in 1991-1992, four on the east and four on the west side. Each chronosequence comprises harvested stands of three age ranges (reference year 1990): regeneration (REG), 3-8 years, immature (IMM) 25-45 years, mature (MAT) 65-85 years, and an old-growth control (OG) >200 years). East-side sites are dominated by Douglas-fir (Pseudotsuga menziesii) with a small component of western hemlock (Tsuga heterophylla) and western red cedar (Thuja plicata). West-side sites are dominated by western hemlock with some amabilis fir (Abies amabilis), western red cedar and Douglas-fir. We previously showed that a single harvesting disturbance had little influence on the organic composition of coarse woody debris and five pools of poorly-decomposed organic matter (Preston et al. 1998; 2002). There was a stronger influence of location, however, with more influence of charcoal residues on the east side vs. woody inputs on the west. We characterized mineral soil C (<2 mm) from 0-10, 10-30 and 30-50 cm depths of the REG and OG plots of one east- and one west- side site. A light fraction (<1.8 g cm-3) was removed by floatation in NaI solution, and the remaining soil was wet-sieved into <53, 53-250, >250 mm particle-size fractions. Preliminary results indicate again, a stronger influence of climate than a single harvesting disturbance. For the east side, most of the C is in the LF and <53 mm fractions, but the west side sites have less C in LF and more in >250 and 53-250 mm fractions. The δC-13 values were lowest for LF, consistent with little-modified plant inputs, while the highest (most enriched) values were found for for the <53 mm fractions from east-side sites. In general, fractions from west-side sites were more depleted in C13 than those from the east sites. Soil C was analyzed by solid-state C-13 CPMAS NMR spectra obtained at 75.47 MHz on a Bruker MSL 300 spectrometer for LF and some HF-treated size fractions, mainly west-side samples with higher C concentration. The LF fraction spectra appear as poorly-decomposed plant material, with some possible char influence. The mineral fractions generally show increasing alkyl/O-alkyl C ratios with depth and decreasing particle size, with little accumulation of aromatic C. The large inputs of lignin from decomposing wood are poorly preserved in mineral-associated fractions, and the highest proportion of OM is preserved as alkyl C. Results from accelerator mass spectrometer (AMS) analysis of C-14 will be incorporated to contribute to understanding C dynamics as affected by climate and a recent harvesting disturbance. Preston, C.M., Trofymow, J.A., Niu, J. and Fyfe C.A. 1998. 13C CPMAS NMR spectroscopy and chemical analysis of coarse woody debris in coastal forests of Vancouver Island. For. Ecol. Manage. 111: 51-68. Preston, C.M., Trofymow, J.A., Niu, J. and Fyfe, C.A. 2002. Harvesting and climate effects on organic matter characteristics in British Columbia coastal forests. J. Environ. Qual. 31: 402-413.