Bacillus-derived alkaline proteases are the major industrial workhorses and the recent trend towards the use of alkaline proteases from these sources in different process applications like detergents, tanning, food, waste treatment and peptide synthesis has increased remarkably becausae of their increased production capacities, high catalytic activity and high degree of substrate specificity (Kumar et al. 1998; Kumar and Takagi 1999). In this paper, we reported a new strain of B. clausii, which produced high levels of an extracellular alkaline protease with optimal pH of 11 and temperature of around 60°C. The protease activity was strongly inhibited by phenylmethylsulphonyl fluoride confirming it as a serine protease (Table 6). We identified that soyabean meal was an effective medium ingredient for the protease production by B. clausii I-52 among the organic nitrogen sources tested (Table 1). With respect to the nitrogen sources, soyabean meal (Glycine max) is one of the potentially useful cost-effective medium substrate because of its easy availability and low-cost as it is produced as a by-product during oil extraction (Gattinger et al. 1990). However, the addition of casein and gelatin showed no or little effect on the protease production in B. clausii I-52. This result was somewhat different from some other Bacillus species. It was earlier reported that the addition of casein substantially improved the protease production in B. licheniformis MIR29 (Ferrero et al. 1996) and Bacillus sp. (Puri et al. 2002). Protease production was increased approx. 30% by the addition of 1% (w/v) casein in B. horikoshii isolated from the haemolymph of a unique Korean polychaeta, Periserrula leucophryna (Joo et al. 2002). The protease yield was greatly enhanced approx. 1·75-fold by the addition of 1% (w/v) wheat flour (Table 3) and, especially, 3·58-fold by the addition of 2·5% (v/v) liquid maltose (Table 4) to a culture medium when compared with a basal medium containing 1·5% (w/v) soyabean meal. However, lactose exhibited a negative effect on the protease production. Contrary to this result, Mabrouk et al. (1999) reported the enhancement of protease production in B. licheniformis ATCC 21415 by the addition of lactose, but a lowered yield was observed with the addition of maltose to the culture medium. Based on the optimization studies, we achieved a yield of 24 270 U ml−1 with specific activity of 28 220 U mg−1 protein when cultivated for 48 h at 37°C in a medium containing (g l−1): soyabean meal, 15; wheat flour, 10; liquid maltose, 25; K2HPO4, 4; Na2HPO4, 1; MgSO4·7H2O, 0·1; Na2CO3, 6 (Table 4). Despite the use of low-cost medium ingredients, the proteases must also exhibit a strong stability against surfactants and oxidants, which have been the common ingredients in modern bleach-based detergent formulations. The protease from B. clausii showed stability and compatibility towards strong anionic surfactants like SDS and oxidizing agents such as H2O2 and sodium perborate. Kobayashi et al. (1995) reported that an alkaline protease from Bacillus sp. KSM-K16 retained approx. 75% activity on treatment with 5% SDS for 4 h. Earlier reports on the stability of alkaline proteases towards oxidants had indicated that an alkaline protease from Bacillus sp. RGR-14 showed 40% loss in enzyme activity with 1% H2O2 (Oberoi et al. 2001), while a subtilisin-like protease from Bacillus sp. KSM-KP43 lost little or no enzyme activity on treatment with 10% H2O2 for 30 min (Saeki et al. 2002). However, there is little published literature available concerning the stability studies of protease towards both SDS and hydrogen peroxide. Gupta et al. (1999) reported that the protease from Bacillus sp. SB5 retained about 60 and 95% of its activity on treatment for 1 h with 1% SDS and 5% H2O2, respectively, while the Bacillus sp. JB-99 protease retained 75 and >95% of its activity on treatment for 1 h with 0·5% SDS and 5% H2O2, respectively (Johnvesly and Naik 2001). Comparing these results, the B. clausii protease exhibited a significant compatibility and stability towards both surfactants and oxidizing agents, which retained its activity of 73 and 116% after incubation for 72 h with 5% SDS and 5% H2O2 (Table 5). As the protease produced by B. clausii I-52 was more stable over a wide range of pH and temperatures and also towards both surfactants and oxidants, it is envisaged that the isolate can be a potential source of alkaline protease for use as additive in industrial applications like detergent industry.