The temperature optimum of enzyme activity, as recently described by Daniel et al. (2001), is thought to play a central role in the adaptation of an enzyme to its environmental temperature. The aim of this research was to exploit this recent development in. determining temperature/activity relationships by correlating the environmental temperature of Antarctic bacteria with the properties of their enzymes. To avoid the problems associated with the comparison of taxonomically distant bacteria, enzymes were sourced from a single genus, Bacillus. Psychrotolerant bacteria were isolated from soils sampled at several sites (Dry Valleys, Bratina Island and Cape Crozier) in the Ross Sea region, Antarctica. At each sampling location, continuous 24 h temperature profiling of the top 2 cm of soil showed exposure to a wide temperature range (up to 20°C difference between maximum and minimum temperature) over the day. Initial characterisation revealed that 26 out of the 52 isolates obtained were Gram positive rods. Random amplified polymorphic DNA (RAPD) analysis of extracted genomic DNA and denaturing gradient gel electrophoresis (DOGE) of the 16S rRNA genes were performed on this subgroup. Five groups of closely related bacteria were identified, along with several unique isolates. 16S rDNA sequencing of eight of these isolates identified them as most likely representing the Bacillus, Planococcus, Arthrobacter, Pseudomonas, Carnobacterium and Psychrobacter genera. Isolates PRl and MV2 (identified as species of Bacillus) were found to have an approximate optimum growth temperature of 27°C and >30°C, respectively. A survey of extracellular proteases, esterases/lipases, glucosidases and galactosidases was initially performed on thermophilic Antarctic bacteria previously isolated from geothermal soils from Mount Erebus, Ross Island. The final choice of an intracellular enzyme, dihydrofolate reductase (DHFR), for temperature characterisation was preferable due to its ubiquitous expression and to allow comparison of the results with thermal properties of DHFRs from other sources. An extrapolation of the temperature dependence of PRl DHFR activity back to zero time revealed a 'real' temperature optima of 44°C, where at higher temperatures in the profile, the enzyme showed a decrease in catalytic rate greater than could be accounted for by irreversible denaturation. A temperature optimum determination for thermophilic DHFR from B. stearothermophilus also displayed a clear peak of activity at 47°C at zero time. Data obtained for the psychrotolerant and thermophilic enzymes were insufficient to confirm the environmental significance of this intrinsic property. However, the idea that 'true' temperature optima reflects the environmental temperature was validated, being lower for the psychrotolerant enzyme compared to the thermophilic enzyme.