Two hemicellulases, cloned from a thermophilic anaerobe (isolate TP8 6.3.3.1) were partially purified and characterised. The ability of these enzymes to hydrolyse the hemicellulose components of various types of pulps and their effects on the bleachability of radiata pine kraft pulp were examined. The xylanase (two preparations) and mannanase had broad pH optima (6-7.5). The xylanases had temperature optima of 70°C (10 min assay) and the mannanase had an optimum of 80°C (10 min assay). All the cloned enzymes were thermostable with more than 95% of the enzyme activity remaining after 24 hours at 70°C, in an unbuffered solution. Buffering affected the thermostability of the enzymes. The mannanase had a half life of 48 min at 85°C. The xylanases were able to hydrolyse xylans but also had a pH-dependent CM cellulase activity, but this was never more than 10% of the xylanase activity. The mannanase hydrolysed mannan, glucomannans and galactomannans. The glucose content of the glucomannans had little or no effect on mannanase activity but the galactose substituents of the galactomannans severely inhibited the hydrolysis of these substrates. The thermostable xylanase hydrolysed more than 75% of soluble radiata pine xylan to low molecular weight (< DP10) xylo-oligomers. These included; xylose, xylobiose, xylotriose, arabinose-substituted xylobiose and xylotriose and 4-O-methyl glucuronic acid substituted xylotrioses. The arabinose substituents were bound to the terminal, non-reducing xylose residue; the glucuronic acid substituents may be bound to the non-reducing or central xylose residues of the substituted xylotriose. The thermostable mannanase also hydrolysed more than 75% of soluble radiata glucomannan, producing mannose, mannobiose, mannotriose, glucomannose and glucomannobiose. The hemicellulases were used to treat (at 70°C, 24 hr) a variety of wood pulps, including kraft (softwood and hardwood), radiata NSAQ pulp, radiata CTMP and radiata TMP. Both enzymes solubilised their respective substrates from these pulps. The chemical pulps were more susceptible to enzymatic attack than the mechanical pulps, possibly due to the greater fibre porosity of the chemical pulps. The xylanase solubilised ~4% of the xylan and the mannanase solubilised 10% of the glucomannan from radiata kraft pulp. In comparison with other xylanases, the thermostable xylanase was less effective in solubilising xylan from kraft pulps. The hemicelluloses of kraft pulps that had been enzyme treated were extracted. The extracted xylan from the xylanase-treated pulp was lower in average DP than the xylan extracted from untreated pulps. The average DP of the extracted glucomannan from the mannanase-treated pulps was no different than that of glucomannans extracted from untreated pulps. The extracted xylan and glucomannan from untreated kraft pulps were readily hydrolysed by the xylanase and the mannanase respectively, suggesting that enzyme inaccessibility to the substrate within the fibre wall was the main cause of their resistance to enzymatic attack. The effect of these enzymes on the bleachability of radiata kraft pulps was studied by measuring the reduction in kappa number (κ), relative to untreated controls, after a standard D/C E prebleaching sequence. The xylanase and the mannanase had little or no effect on the bleachability of radiata kraft pulps. Xylanase treated kraft pulp required a protease treatment after the hemicellulose treatment but before bleaching, in order to elicit any improvement in bleachability (~7% drop in κ relative to untreated control pulps). Mannanase treated kraft pulp showed no improvement in bleachability. For comparison commercial xylanases (including a thermostable xylanase) were used to treat radiata kraft pulps and did improve bleachability significantly. These enzymes also solubilised three to four times more xylan than the cloned, thermostable xylanase. Therefore, the ineffectiveness of the cloned xylanase may be due to its very limited ability to hydrolyse intra-fibril xylan in kraft pulp fibres.