In this note an analysis of USSA 1976 has been made about its physical assumptions and atmospheric tables.The main results are:1.Temperature profile T(Z) may be defined with absolute temperature in full height range without using molecular-scale temperature T
M. 2.The elliptical segment of T(Z) will give a sudden heat sink of the order of 10
-7J/m
3.s at 110 km which is too large.3.It seems the only systematic mistake in atmospheric tables is the coefficient of thermal conductivity.4.Adetailed discussion is given about the possible choices to connect the computations of the homosphere and the upper heterosphere.5.We suggert to introduce CO
2 as a component in the upper atmosphere model and lower the computing interface down to 80 km.