Qantitative analysis of the interrelation of "big" and "little" science on the example of Research and Development of higher education in the USA has been made. The difference in the growth rates of "big" and "little" science is explained with the help of scientometric index of capital expenditures per researcher. An attempt has been made to compare the dynamics of efficiency of "big" and "little" science on the base of mean duration of a research project. Possibilities of an alternative index of a relative amount of preliminary researches (preprojects) are pointed out. © 1987 Akadémiai Kiadó.