Concerning fresh and acetone-dried cells of amylase-producing bacteria, Bac. subtilis var. amyloliquefaciens Fukumoto, phosphorus distribution was determined by Schmidt-Thanhauser method. It was estimated that DNA-P is very low in content amounting to about 0.02 % of fresh cell or about 0.19 % of dried cell, while much more RNA-P is contained by 10 to 15 times of DNA-P. Therefore, it was suggested that this bacterial cell can not be suitable to DNA source in general, even if cells were multiplied manufacturally and discarded without any utilization. Using acetone-dried cells, several procedures for preparation of bacterial DNA were discussed by testing them. Firstly, cells were disrupted by grinding with some aid in mortar and extracted preliminarily with diluted NaC1 solution or 0.2 % citric acid to remove non-DNA substances as far as possible. From cell mass remained, DNA was extracted with 2 M NaCl and purified by several steps of treatment. These are mainly consist of Sevag's method to remove proteins by chloroform-protein gel formation in concentrated NaC1 solution, and of Chargaff's procedure to separate DNA from RNA by fractionation with alcohol in 10 % CaCl_2 solution. DNA was finally precipitated by adding it into 2 volumes of alcohol, removed moisture by stepwise treatment with alcohol of increasing concentration, washed with ether and dried in a vacuum desiccator, becoming white powder. However, it was found that this procedure is not satisfactory to preparation of DNA, regarding both purity and yield. Secondly, Dounce's method using detergent was investigat e d. In this case, Japanese product, Emal OD, was used instead of original one, Duponol. However, DNA preparation also has much inpurities and its yield was very low. Finally, DNA was extracted with 0.5 % phenol according to Braun except that preliminary treatment with chloroform freezing cells had been omitted. From the extract, white powder of DNA was obtained after similar purification. The method was seemed to fit more or less for the purpose. However, the combination of phenol extraction with grinding in mortar was most suitable among all procedures tested. In connection with these trials, some properties of bacterial nucleoprotein as well as nucleic acid were observed. Especially, it may be noticed that Sevag's method of removing protein is not complete. However, remaining protein can be eliminated by further treatment with chloroform in concentrated NaCl solution, when DNA precipitate had been fractionated with alcohol in CaC12 solution by Chargaff's method after completion of Sevag's procedure. Furthermore, it must be remembered that, in the procedure of Chargaff, RNA is precipitated together with DNA in CaC12 solution by adding 0.3 volumes of alcohol, in contrast from those of yeast. At any rate, DNA preparation from amylase-producing bacteria is very difficult and more suitable procedure must be investigated in respect of both purity and yield.