Chemical senses play a critical role in predator-prey and social interactions of many animals. Predators often evoke adaptive escape responses by prey, one of which is the release of chemicals that induce adaptive avoidance behaviors from both predators and conspecifics. I explore the use of chemicals in predator-prey and social interactions, using a crustacean model system, the spiny lobster. As predators, spiny lobsters are opportunistic, polyphagous feeders, and they rely heavily on their chemical senses during feeding. Some of their potential prey deter attacks through chemical defenses that act through the spiny lobsters’ chemical senses. An example of this is sea hares, Aplysia californica, which secrete an ink when vigorously attacked by sympatric spiny lobsters, Panulirus interruptus. I show that that this ink defends sea hares from spiny lobsters through several mechanisms that include phagomimicry, sensory disruption, and deterrence, and that the ink’s efficacy is enhanced by its naturally high acidity. As prey, spiny lobsters rely heavily on their chemical senses to assess risk from predators. One way to assess risk of predation is through ‘alarm cues’, which are injury-related chemicals. I show that injured Caribbean spiny lobsters, Panulirus argus, release alarm cues in their hemolymph, and that nearby conspecifics detect these cues using olfaction. Hemolymph from conspecifics induces primarily alarm behavior in the form of retreat, sheltering, and suppression of appetitive responses. In contrast, hemolymph from heterospecifics, depending on phylogenetic relatedness, induces either mixed alarm and appetitive behaviors or primarily appetitive behaviors. Spiny lobsters also use chemical cues to assess risk during social interactions with conspecific. I show that spiny lobsters use urine-borne chemical signals and agonistic behaviors to communicate social status and that these chemical signals are detected exclusively by the olfactory pathway. Dominant animals increase urine release during social interactions, whereas subordinates do not. Experimental prevention of urine release during interactions causes an increase in agonism, but this increase is abolished when urine of dominants is reintroduced. My findings lay the foundation for neuroethological studies of risk-assessment systems mediated by intraspecific chemical cues.