The factors that influence the distribution and chemical composition of the main soil series in Malawi were reviewed and five widely cultivated soil series were selected for study. Two soil samples, one virgin and another cultivated, were taken from each series. After establishing the applicability of the ascorbic acid reduction procedure for phosphorus determination in various soil extracts, the soils were analysed for physicochemical properties, phosphate retention capacity, total and organic phosphorus; and for inorganic phosphorus soluble in the following solutions: 0.5M acetic acid (HOAc); acetic acid + sodium acetate buffer (HOAc + NaOAc) at pH H.8; 0.5M NaHC03 at pH 8.5; 0.01M CaCl2 (from which the equilibrium phosphate potential I0 and phosphate ion activity aH2P0it were calculated); 0.03M NII. F + 0.025M HC1 (Bray-Pi);0.03M NH. F + 0.01M H2SO4 (Maida-P); 0.05M HC1 + 0.025M H2SOit (Nelson-P);cold 0.1M NaOH; and 1 per cent w/v citric acid (H3Ci). A.modified version of the Chang and Jackson procedure was used to fractionate the inorganic soil phosphorus fractions and an anion-exchange resin method developed by the author using "Amberlite" resin IRA-UlO, Cl form, was also applied. All the soils were acid and low in general piant-nutrient status, but varied from coarse to medium fine texturally. About U9 per cent of the mean total phosphorus was organic and most of the inorganic phosphorus was in a sesquioxidic-bound form; the latter was attributed to the excessive weathering processes these soils have undergone. Total phosphorus contents were higher in the cultivated than in the virgin soils, and the observed differences were statistically significant. The Stanford and DeMent seedling bioassay technique was studied and modified. The modified technique using oats (Ayr Commando variety) was effective for determining soil P status, and so it was used to discover which soil chemical test most effectively described the soil phosphate supplying ability. Correlation coefficients and stepwise regressions of P uptake on chemical P tests were computed as an aid in this evaluation. Phosphate retention capacity values were highly related to contents of organic matter, clay, iron, aluminium and titanium. The total variation in the P values soluble in all the solutions except HOAc +/&QAc and 0.01M CaCl2 (l0) was accounted for chiefly by aluminium-bound P. The variation in HOAc H-Z/oOAc or ID was mainly ascribable to saloid-bound phosphorus. When each of the chemical soil P tests was regressed on saloid-bound, aluminium-bound, iron-bound, and calcium-bound phosphate fractions by the stepwise regression method, complications arising from what was postulated to be omitted variables and/or near collinearity of the independent variables made interpretations of the regression coefficients difficult. Possible modifications have been suggested and Dr. M. N. Court, the author's supervisor, hopes to write new computer programs for further analysis of the data. The amounts of P extracted chemically were significantly interrelated and each-of them, except HOAc + NaOAc, was significantly related to P uptake. In order of decreasing precision for predicting P uptake by the oat seedlings, the laboratory methods were as follows: 0.5M NE1HCO3 - anion-exchange resin - Nelson-P > Bray-P. > HsCi > HOAc > Maida-P " 0.1M NaOH > aH2P0i+> ID >> HOAc + NaOAc. The 0.5M NaHC03~P, resin-P, and Nelson-P did not differ very much in their predictive values; they accounted for, respectively, 86, 85 and 8U per cent of the total P uptake variability. The main source of plant-available P was aluminium-bound P. Little or no improvement in predictive value was obtained from inclusion of organic P in a regression of P uptake on each of the soil P tests . : To investigate if mineralization of organic P would be conducive to an increase in the total P uptake, soil samples moistened to TO per cent moisture equivalent were incubated at 30°C for 2 weeks before using them for P absorption studies.