Although females rarely experience strong mate limitation, delays or lifelong problems of mate acquisition are detrimental to female fitness. In systems where males search for females via pheromone plumes, it is often difficult to assess whether female signaling is costly. Direct costs include the energetics of pheromone synthesis and emission, and attention from unwanted eavesdroppers such as parasites, parasitoids and predators. Suboptimal outcomes are also possible from too many or too few mating events, or near simultaneous arrival of males who make unwanted mating attempts (even if successfully thwarted). We show that, in theory, even small costs can lead to a scenario where young females signal less intensely (lower pheromone concentration and/or shorter time spent signaling) and only increase signaling effort as they age and gather evidence (whilst still virgin) on whether sperm limitation threatens their reproductive success. Our synthesis of the empirical data available on Lepidoptera supports this prediction for one frequently reported component of signaling — time spent calling (often reported as the time of onset of calling at night) — but not for another, pheromone titer. This difference is explicable under the plausible but currently untested assumption that signaling earlier than other females each night is a more reliable way of increasing the probability of acquiring at least one mate than producing a more concentrated pheromone plume.