According to the environmental cues hypothesis, female birds use information available to them in the early-season environment to fine-tune egg production annually. However, support for the hypothesis derives largely from correlational studies. In each year from 2002 to 2006, which spanned a period of extreme variation in environmental conditions, we removed eggs from early-laying rhinoceros auklets Cerorhinca monocerata, burrow-nesting seabirds that lay a single-egg clutch. We then measured their relaying rates, relaying intervals, and breeding success. We also monitored the timing and success of breeding in control pairs, and control chick diets. If the experimental females base their relaying decision on early-season cues, then we predict that few will relay in years in which early-laying control birds breed unsuccessfully, and in which a preferred prey species, Pacific sandlance Ammodytes hexapterus, is in short supply in nestling diets. Results matched neither prediction. In each year, almost all (88–90%) of the experimental females relaid, despite that the control pairs’ breeding success (32–87% fledged chicks), and their chicks’ diets (twofold variation in proportion of sandlance), varied markedly. We conclude that female rhinoceros auklets did not modify their relaying decision in response to variation in environmental conditions, although relaying intervals and their own breeding success (0–78%) covaried negatively. Our results may have important implications related to using seabirds as monitors of the marine environment.