Crayfish have well developed sense organs, but clear-cut relationships with their behavioural use have hardly been established. We studied the possible use of vision, olfaction, and touch during agonistic behaviour, assuming that the outcome of agonistic interactions primarily depends on the input from one of these sensory organs. Agonistic encounters were studied in triads of crayfish, intact and with reversible blockage of vision, olfaction, or touch. Tension actions (threat, strike, fight, or avoidance) categorized as positive or negative depending on the crayfish initiating them, allowed us to identify the dominant and submissive animals. The contribution of sense organs to the outcome of interactions was tested during agonistic behaviour by blocking them (one at a time) after the establishment of a hierarchy (after-experiments) of before it (before-experiments). Under control conditions, a large number of contacts allowed animals to establish a dominance order on the first day of agonistic interactions, and the number of positive contacts between animals diminished in subsequent days. Visual or chemical blockage in after-experiments did not change the dominance order, but positive contacts decreased or increased, respectively. Blinded-before animals established a dominance order in the first 3 days of agonistic interactions showing an elevated number of positive contacts during the observation period. A similar result occurred in anosmic-before animals. Results from crayfish in which antennae were immobilized were similar to those from controls. Results suggest that at least two sensory modalities are necessary to gather information about conspecifics. Once the order is established (learned) any one of the senses is sufficient to maintain it. We speculate that if a chemical compound is involved in the maintenance of the dominance order, it is released after localization of a conspecific by vision or touch, a manoeuvre that could minimize expenditure of a costly resource.