Functional response experiments were performed in the laboratory to examine the effect of prey density (as observed in the field) on feeding behaviour, and to measure handling-times and attack-rates for each instar and adult of Ranatra dispar Montandon (Heteroptera: Nepidae) feeding on five size-classes of its common prey, Anisops deanei Brooks (Heteroptera: Notonectidae). The most generally applicable response was the Type 2, although for both the predator fifth instar and adult female and male feeding on the two smallest prey sizes, the asymptote or plateau was not observed even at the highest prey density given. Generally, the handling-time increased as prey-size increased, and decreased as the predator size increased. The attack-rate surface was far more complex. For the first two predator instars (I and II), the maximum attack-rate occurred on the smallest prey sizes (1 and 2). The maximum attack-rate for predator instar III was almost the same for prey sizes 1 and 2, that of predator instar IV was greater for prey size 2, while in the three largest predator sizes (V, female and male), the maximum attack-rate was found for prey size 3. Predator instar V had the largest attack-rate values over all prey sizes, and both the predator adult female and male had lower attack-rates for various prey sizes than instars V, IV and, to some degree, III. The results support the suggestion that small predator instars will usually compete with large instars for prey, unless they are spatially or temporally separated. Observations in the field indicate that a distinct age-specific spatial distribution exists in R. dispar and the prey, A. deanei, with the smallest individuals being found predominantly in the shallow (littoral zone) water, while the larger individuals are found in the deeper water.