This thesis presents an investigation of carbonyl sulphide as a new fumigant and related methodology studies. The first part involved the investigation of a new fumigant - carbonyl sulphide, which has the potential to replace methyl bromide. Its biological response or activity was investigated, e.g. toxicity to target organisms and phytotoxicity, environmental and worker safety considerations. In the second investigation, analytical methods were developed for the determination of fumigant movement through timber and fumigant residues in grains as well as a method of chemical fractionation to determine the fate of carbonyl sulphide. A comprehensive literature review of 161 references in these two areas is reported. Carbonyl sulphide was highly toxic to adults of three coleopteran species tested, namely Rhyzopertha dominica (F.), Tribolium confusum du Val, and Sitophilus oryzae (L.), the most sensitive species was R. dominica. For 6 hr exposure at 25�C, the L(CxT)95 value for R. dominica, S. oryzae and T. confusum were, respectively, 36.48, 99.82 and 113.0mg h L-1. Carbonyl sulphide inhibited 100% of mould in wet wheat and more than 90% of mould on dry wheat at lOOmg L-1. Both carbonyl sulphide and hydrogen cyanide were low in phytotoxicity without affecting germination of wheat, at levels needed to control insects. Unlike hydrogen cyanide, carbonyl sulphide can be used at minimum levels without decreasing plumule length of wheat. Chemical data on the sorption of carbonyl sulphide are compared with data from methyl bromide. The levels of carbonyl sulphide in the headspace of five commodities (wheat, barley, paddy, sorghum and peanut) and timbers (hardwood and softwood) decay more slowly than do levels of methyl bromide. Carbonyl sulphide was blown through a column of wheat as easily as was phosphine and more easily than was methyl bromide, and its front was blown out faster than phosphine and methyl bromide. Movement of two fumigants (methyl bromide and carbonyl sulphide) through, and sorption on, softwood and hardwood were studied. Each fumigant was sorbed less on softwood than on hardwood and penetrated softwood better than hardwood. Carbonyl sulphide penetrated timber better than did methyl bromide, and was less sorbed on timber. A rapid method of solvent extraction was developed to enable rapid estimation of the amount on intact fumigant sorbed in wood. This procedure enabled near quantitative recovery of methyl bromide as either intact fumigant or as bromide ion. Carbonyl sulphide residue in unfumigated wheat was found to be around 25- SOppb. Carbonyl sulphide left little residue on fumigated grains. Desorption of carbonyl sulphide from the wheat was extremely fast, 85% of it was released after one day aeration which was very much greater than that of methyl bromide and carbon disulphide. After 6 days aeration the incorporation of 14COS on mungbean, wheat, paddy, rice and safflower was lower than 7Oppb (calculated as COS equivalent). Food value or nutritional quality of foodstuffs is not harmed by carbonyl sulphide fumigation. This result was assessed by identifying any nonreversible change or combined residues in biochemical fractions of commodities including lipids, protein, amino acids, carbohydrate, etc., and no irreversible reaction between carbonyl sulphide and any constituent such as B vitamin, atocopherol, lysine, maltose and starch. Fumigants did not affect lipids, although each fumigant was applied to wheat at exaggerated concentrations, nor wheat germ oil and canola oil treated with extremely high concentration of fumigants. Factors which affect analysis of fumigants including stability of chemicals in extraction solvent and partitioning of fumigant between solvent and air, were examined. The partition ratio, defined as the fumigant concentration in extraction solvent to that in the headspace, varied with fumigant. Methods for multi-fumigant analysis were developed or modified and gave high recoveries and efficiency. The procedure of Daft of solvent extraction followed by partitioning was modified by being performed in sealed flasks. This raised the recovery of carbonyl sulphide, methyl bromide, phosphine and carbon disulphide. Recoveries were near quantitative at levels down to 6-16ppb (w/w) for tested fumigants. Thus the modified Daft method can be adapted to enable determination of the main fumigants used on staple foodstuffs. Microwave irradiation method give higher efficiency of removal of fumigants from grains. Limits of quantification were < 0.2ng g-1 (ppb w/w) for each tested fumigant. The detection limit of COS was calculated, as natural levels of the fumigant were detected in commodities. These are feasible, simple and rapid (< 2 min.) to be use to analyse fumigant residue in grains. Carbonyl sulphide has potential as a fumigant for grain and timber and may replace methyl bromide in some uses, subject to further investigation in commercial situations.