This paper summarizes the recent work on fast neutron dosimetry done by the Health Physics Division of the Oak Ridge National Laboratory. Neutrons do not ionize directly, but by various reactions with tissue they produce particles which do ionize. In the case of fast neutrons the important reactions are the elastic collisions with hydrogen, carbon, nitrogen, and oxygen atoms. The recoil nuclei are charged and spend the energy gained from the neutron by ionization and excitation of the tissue atoms. Thus, it is important to measure the amount of energy imparted to a gram of tissue. Accordingly the unit accepted for measuring the tissue dose of fast neutrons is the roentgen-equivalent-physical (rep), which is the amount of radiation needed to dissipate 93 ergs per gram of soft tissue. The unit as defined cannot be measured directly in practice. However, the amount of ionization produced by the radiation is a fairly accurate index of the amount of energy used in exciting and ionizing tissue. We wish to discuss two methods that have been developed specifically for measuring the dose due to fast neutrons in the presence of gamma radiation. Count-Rate Dosimeter The first principle is the fact that a counter can be so designed that its count rate dependence on energy is the same as the variation of the dose with energy. The amount of energy that is given to the hydrogen, carbon, nitrogen, and oxygen atoms in a gram of tissue can be calculated if the various cross sections are known, and if one assumes that the scattering of all the atoms is isotropic in the center of mass system. The result of such a first collision calculation is shown in Figure 1. The details of a detector which has a dependence on neutron energy that approximates this curve have been described (1). We will now indicate how the response of such a detector can be calculated. Figure 2 is a schematic drawing of a proton radiator which may be a plane slab of paraffin or polyethylene. The number of protons dN contributed from dt as a result of elastic collisions is where A is a constant, σ(E) is the hydrogen cross section as a function of the neutron energy E, and the quantity η(t, E, B) is the fraction of those protons produced in dt which reach the surface of the slab with an energy equal to or greater than B, the bias energy. If one assumes that the range of protons in the slab is expressed as kEn, where k and n are constants, and if scattering is isotropic in the center of mass system, then η can be computed. The integration of equation (1) over all dt then gives for E ≤ E0, and for E ≥ E0n and E0 》 B, where E0 is energy of the proton having a range t0.