Aim: Statistical optimization for maximum production of a hyperthermostable, Ca2+- independent and high maltose-forming alpha-amylase by Geobacillus thermoleovorans. Methods and Results: G. thermoleovorans was cultivated in 250 ml flasks containing 50 ml of chemically defined glucose-arginine medium ( g l(-1): glucose 20; arginine 1.2; riboflavin 150 mug ml(-1); MgSO4. 7H(2)O 0.2; NaCl 1.0; pH 7.0). The medium was inoculated with 5 h-old bacterial inoculum (1.8 x 10(8) CFU ml(-1)), and incubated in an incubator shaker at 70 degreesC for 12 h at 200 rev min(-1). The fermentation variables optimized by 'one variable at a time' approach were further optimized by response surface methodology (RSM). The statistical model was obtained using central composite design ( CCD) with three variables: glucose, riboflavin and inoculum density. An over all 24 and 70% increase in enzyme production was attained in shake flasks and fermenter because of optimization by RSM, respectively. A good coverage of interactions could also be explained by RSM. The end products of the action of alpha-amylase on starch were maltose (62%), maltotriose (31%) and malto-oligosaccharides (7%). Conclusions: RSM allowed optimization of medium components and cultural parameters for attaining high yields of alpha-amylase, and further, a good coverage of interactions could be explained. The yield of maltose was higher than maltotriose and malto-oligosaccharides in the starch hydrolysate. Significance and Impact of the Study: By applying RSM, critical fermentation variables were optimized rapidly. The starch hydrolysate contained a high proportion of maltose, and therefore, the enzyme can find application in starch saccharification process for the manufacture of high maltose syrups. The use of this enzyme in starch saccharification eliminates the addition of Ca2+.