During the dying phase, patients often receive interventions that are not aimed at promoting their comfort. We investigated how recognition of the dying phase affects the use of interventions by comparing patients for whom the dying phase had been recognized with patients for whom it had not been recognized. We included 489 of 613 patients (80%) who died either in a hospital, nursing home, or primary care setting between November 2003 and February 2006. After the death of patients, nurses filled in questionnaires, and patient records were searched for information about therapeutic and diagnostic interventions applied during the dying phase. Caregivers had recognized the dying phase of 380 patients (78%). The number of patients who had received diagnostic interventions during the last three days of life was significantly lower when the dying phase had been recognized (39% vs. 57%) (p = 0.00). Therapeutic interventions were used in similar frequencies in both groups. We conclude that recognition of the dying phase reduces the number of undesirable diagnostic interventions.