It is generally assumed that chemical signals (sexual pheromones) constitute the primary stimulus for sexual attraction in many mammals. However, it is unclear whether these pheromones are volatile or nonvolatile and which sensory systems are involved in their detection (vomeronasal and/or olfactory). Moreover, it has been demonstrated that experience influences the behavioral response to sexual pheromones and the sensory systems implicated. In order to clarify this issue, the attractive properties of volatile and nonvolatile components of the male-soiled bedding have been analyzed in female mice that had no previous experience with adult male-derived chemical signals (chemically naïve females) using two-choice preference tests. The results indicate that some nonvolatile male-derived substances exert an innate attraction to females, but volatiles derived from male-soiled bedding do not attract chemically nai;ve females. Therefore, the primary attractive sexual pheromone includes a nonvolatile compound (e.g. major urinary proteins, MUPs). On the other hand, male-derived volatiles become attractive to females because of repeated exposure to male-soiled bedding. This represents a Pavlovian-like associative learning in which previously neutral volatiles (very likely odorants) acquire attractive properties by association with the nonvolatile, innately attractive pheromone(s). These findings indicate that not only the sexual but also the 'chemical' experience (previous experience with sexual pheromones) has to be taken into account to interpret the role of chemicals as releaser or primer pheromones. The sensory systems involved in the detection of these stimuli and the neural basis of the odor-pheromone association are discussed.