The use of a direct reactivity meter for control-rod calibration was studied. The reactor model was simplified by reducing the number of delayed-neutron and photoneutron groups from 15 to 6, and by putting dn/dt equal to zero, without greatly affecting the accuracy of the reactivity measurements. The influence of errors in the knowledge of the parameters of the remaining six delayed-neutron groups was studied. Measurements were made on two reactors: ECO, a cold reactor; and ISPRA-I, a reactor with a strong long-lived photoneutron source. The measurements performed with the direct reactivity meter were compared to normal rod-drop measurement techniques and period measurement techniques, and were found to agree in general to ±2%. The long-lived source term, which depends on the reactor operation history, was found to a precision of better than ±1%.