Many studies have been conducted on K fixation by soils and pure minerals. Much of the previous work has focused on K fixation and release by the clay size fraction (< 2 μm). However, in soils containing considerable amounts of silt- and sand-size K-fixing minerals, such as vermiculite and hydrobiotite, potassium fixation capacity may be substantial in the coarser size fractions. We hypothesized that the alteration of biotite to vermiculite in silt and sand fractions in soils derived from granitic alluvium would result in high K fixation capacity in those fractions, while charge reduction in the vermiculite, or alteration to smectite, would result in lower K fixation potential in the clay fraction. To test our hypothesis, we measured the contribution of clay, silt, very fine and fine sand fractions to K fixation in five soils derived from granitic Sierra Nevada alluvium in the San Joaquin Valley (SJV), California. We sampled soils from Buttonwillow (Vertic Torrifluvent), Armona (Fluvaquentic Endoaquoll), Gepford (Typic Natraquert), and two Kimberlina (Typic Torriorthents) map units. Soil samples were collected by horizons to a maximum depth of 120 cm. The sites received no K fertilizer in the year prior to sampling. Sodium-dispersed soil samples were separated into clay, silt, and sand fractions by settling, centrifugation, and sieving. Mineralogical composition of size fractions was determined by X-ray diffraction. We measured K fixation capacity of the whole soil and of the size fractions as follows: Samples were shaken for 1 h with 20 mmol K kg^-1, followed by a 30-min extraction with NH₄Cl. The K not displaced by NH₄ was considered to be fixed and was calculated as the difference between the initial K added and the extracted K. Clay fractions were freeze-dried, and silt, Very Fine Sand (VFS), and fine Sand (VS) were air-dried prior to K fixation procedures. Results from this method were highly correlated (r² = 0.95) with results from a seven-day incubation procedure (Cassman et al., 1990). The soils ranged in texture from clay (maximum of 56% clay) to sandy loam (minimum of 5% clay). Smectite dominated the clay fractions of Buttonwillow, Armona and Gepford soils. Mica was a major clay-fraction component of both Kimberlina soils. Silt fractions of all soils were dominated by vermiculite and hydrobiotite. Quartz, amphiboles, mica, feldspars and some vermiculite were present in the VFS and FS fractions of all soils. Soil materials from most of the soil horizons fixed K, and values ranged from 1.6 to 18.9 mmol K kg^-1, comprising about 8 to 95 % of the K added. In all pedons, K fixation increased with depth. Armona loam fixed the most K (Depth-Weighted Average (DWA) 15.5 mmol K kg^-1) and Gepford clay the least (DWA of 4.5 mmol K kg^-1). Overall, DWA K fixation by clays for all pedons was -4.33 mmol K kg^-1 of fraction (or -0.92 mmol kg^-1 of soil), meaning that K was released, not fixed, by the <2μm fraction. Calculated on a fraction basis, K fixation was highest in the silt fractions in most of the pedons, but in subsurface horizons of the Kimberlina and Armona soils, FS and VFS fixed more K than silt fractions (8.4 to 18.5 mmol kg^-1 of FS and 7.4 to 15.8 mmol kg^-1 of VFS vs. 6.2 to 13.8 mmol kg^-1 silt). Calculated on a whole soil basis, the contribution to K fixation capacities from the various size fractions was slightly different than calculated on a fraction basis and depended on the relative proportions of the size fractions. In summary, in SJV soils derived from Sierra Nevada granitic alluvium, most of the clay fractions were dominated by smectite and did not fix K. The highest percentage of added K was fixed by the silt fractions, which are dominated by vermiculite and hydrobiotite. Very fine and fine sands also contributed significantly to K fixation, and in some pedons where these fractions are dominant, they constituted the main locus of K fixation. Cassman, K.G., D.C. Bryant, and B.A. Roberts. 1990. Comparison of soil test methods for predicting cotton response to soil and fertilizer potassium on potassium fixing soils. Commun. Soil Sci. Plant Anal. 21:1727-1743.