Pulsed-neutron experiments have been performed in a borehole model to determine the effects of tool position, borehole fluid, and source-detector spacing on pulsed-neutron/thermal-neutron logging. Neutron-time distributions were measured with various combinations of the above parameters in a sand model of 32.5% porosity filled with fresh water or salt water with 230 g/liter NaCl. Neutron lifetimes determined from the distributions indicate the degree of validity of the assertion that undesirable borehole effects are largely eliminated by this logging method. A numerical computer code (CUNLAP) has been developed to solve the time-dependent, three-group diffusion equations which apply to borehole geometry. Results of test calculations are presented and compared to the experiments in a semiquantitative way. The numerical results are also compared to those of an analytical, fundamental-mode calculation. It is shown that the latter approach is inappropriate for the type of measurement and size of system used in experiments of this nature.