In order to investigate the development of organizational identification during a merger, a quasi-experimental case study was conducted on a pending merger of police organizations. The research was conducted among employees who would be directly involved in the merger and among indirectly involved employees. In contrast to earlier studies, organizational identification was measured as the expected identification prior to the merger. Five determinants were used to explain the employees' expected identification: (a) identification with the pre-merger organization, (b) sense of continuity, (c) expected utility of the merger, (d) communication climate before the merger and (e) communication about the merger. The five determinants appeared to explain a considerable proportion of the variance of expected organizational identification. Results suggest that in order to obtain a strong identification with the soon-to-be-merged organization, managers should pay extra attention to current departments with weaker social bonds as these are expected to identify the least with the new organization. The role of the communication variables differed between the two employee groups: communication about the merger only contributed to the organizational identification of directly involved employees; and communication climate only affected the identification of indirectly involved employees.