An analytical neutron sensor response model and methods for measurement of neutron sensor (compensated ionization chamber) transient response have been developed and evaluated. In situ measurement methods to meet the provisions of the Instrument Society of America Draft Standard dS67.06, Institute of Electrical and Electronics Engineers 338-1977, and U.S. Nuclear Regulatory Commission Guide 1.118 are included. In one in situ method, the high-voltage sensor power supply is perturbed and subsequent sensor response measured. The response is analytically and experimentally related to the response of the sensor to a transient change in radiation flux. Random signal analysis was a second in situ technique evaluated to monitor the transient response of the neutron sensor. In this method the power spectrum of the inherent random fluctuations from the neutron sensor output is measured and analyzed. Transient response was experimentally and analytically evaluated to identify mechanisms that may cause degradation in the response of neutron sensors. Response time degradation was investigated by changing the sensor and signal cable response time in a predictable manner (through changes in the detector fill gas and the use of a delay line and different terminations in series or parallel with the signal cable). Sensors and attached cables having different response times were evaluated using power supply perturbation, transient change in radiation flux, and analysis of the random signals from the neutron sensor. The primary objectives of the experimental evaluation were to correlate the measured response time using transient radiation flux changes with response to a power supply perturbation and to confirm the analytical model. The primary objectives of developing the analytical model of sensor response were to predict response time and to evaluate degradation mechanisms. It is shown that degradation in neutron sensor response time, which may not be significant to the operation of a reactor protection system, is related to degradation in sensitivity and linearity, and that simulated degradation in response time can be detected through the two techniques developed.