This paper focuses on the fluid communication and injection responses in the Pekisko B Pool between January 1990 and September 1997. The analysis was based on responses of producers to changes in the injection rates. The strength of oil response is measured as the correlation between the oil rate and the injection rate. In a similar fashion, the strength of the total fluid responses is measured as the correlation between the total fluid rate and the injection rate. These responses and the corresponding time lags are presented in the form of special XY-diagrams. One form of these diagrams shows normalised responses of wells around each injector. The second form shows the long distance communication between injectors and producers that are characterised by strong responses above the predefined level. Results indicate very good short and long distance communication between injectors and producers. Waterflood responses can be segmented into three geographical areas: northern, central and southern. The oil responses correlate strongly to the injection in the central part of the pool. Weaker responses and longer time lag response was observed in the southern pool area, and strong interference between injectors was detected in the northern pool area. A more detailed analysis, involving pressure data, geology and completion intervals, demonstrate that the existing waterflood inefficiencies are related to inadequate injection volumes in the southern pool area and poor injection well distribution in the northern part of the pool. The study proved that the applied methodology is very efficient in diagnosing the waterflood performance and helping to optimise waterflood design. This analysis very quickly identified communication or lack there of, which had only partially been understood through lengthy geological study of the pool. This method should be used in the early stages of study to very quickly "zero in" on areas of difficulty due to complicated configuration or geological/depositional complexities.