Industrial strain improvement is concerned with developing or modifying microorganisms used in production of commercially important fermentation products. The aim is to reduce the production cost by improving productivity of a strain and manipulating specific characteristics such as the ability to utilize cheaper raw materials or resist bacteriophages. The traditional empirical approach to strain improvement is mutation combined with selection and breeding techniques. It is still used by us to improve the productivity of organisms in amino acids, organic acids and enzymes production. The breeding of high L-lysine-producing strain Au112 is one of the outstanding examples of this approach. It is a homoserine auxotroph with AEC, TA double metabolic analogue resistant markers. The yield reaches 100 g/l. Besides, the citric acid-producing organism Aspergillus niger, Co827, its productivity reaches the advanced level in the world, is also the result of a series mutations especially with $^60Co{gamma}$-radiation. The thermostable .alpha.-amylase producing strain A 4041 is the third example. By combining physical and chemical mutations, the strain A 4041 becomes an asporogenous, catabolite derepressed mutant with rifamycin resistant and methionine, arginine auxotroph markers. The .alpha.-amylase activity reaches 200 units/ml. The fourth successful example of mutation in strain improvement is the glucoamylase-producing strain Aspergillus niger SP56, its enzyme activity is 20,000 units/ml, 4 times of that of the parental strain UV-11. Recently, recombinant DNA approach provides a worthwhile alternative strategy to industrial strain improvement. This technique had been used by us to increase the thermostable .alpha.-amylase production and on some genetic researches.