The dependency at discharge instrument, developed to evaluate the needs for care of hospitalized patients on the day of discharge, was tested for its usefulness in measuring resource consumption in home health care. The instrument evaluates dependency in relation to bathing/hygiene, activity, technical procedures, and monitoring signs and symptoms. The sample used in this study consisted of 150 subjects admitted to one of three types of home health care agencies. Interrater reliability and internal consistency of the instrument were high. Factor analysis produced results similar to those of the developers for three of the four scale items. The item concerning need for assistance with procedures had a much lower loading, indicating that it had little in common with the other items. Criterion-based validity was measured by testing the instrument's ability to predict the number of registered nurse and home health aide visits, and length of enrollment. Findings indicated that the instrument had a modest level of criterion-based validity in predicting the use of registered nurse and home health aide services for the hospital-affiliated agency, but was relatively ineffective for public health and for-profit agencies.