In many animal societies, subordinates exhibit down-regulated reproductive endocrine axes relative to those of dominants, but whether this 'physiological suppression' arises from active interference by dominants or subordinate self-restraint is a matter of debate. Here we investigate the roles that these processes play in precipitating physiological suppression among subordinate female meerkats, Suricata suricatta. We show that, while subordinate females are known to suffer stress-related physiological suppression during periodic temporary evictions by the dominant female, their low estrogen levels while within their groups cannot be readily attributed to chronic stress, as their fecal glucocorticoid metabolite levels during this time are comparable to those of dominants. The low estrogen levels of subordinate females also cannot be explained simply by self-restraint due to factors that could reduce their payoff from maintaining their fertility regardless of the presence of the dominant female (young age, a lack of unrelated mates, poor body condition and limited breeding experience), as substantial rank-related differences in fecal total-estrogen metabolite levels remain when such factors are controlled. We suggest that this residual difference in estrogen levels may reflect a degree of subordinate restraint due in part to the dominant female's ability to kill their young. Accordingly, subordinate female estrogen levels vary in association with temporal variation in the likelihood of infanticide by the dominant. Attempts to identify the causes of physiological suppression should be cautious if rejecting any role for dominant interference in favor of subordinate restraint, as the dominant's capacity to interfere may often be the reason why subordinates exercise restraint.