How does the content of secrets relate to their harms? We identified a data-driven model (across five empirical steps), which suggested that secrets are generally seen to differ in how immoral, relational, and profession/goal-oriented they are (Study 1). The more a secret was consensually perceived to be immoral, relational, and profession/goal-oriented, the more that secret was reported to evoke feelings of shame, social connectedness, and insight into the secret, respectively. These three experiences independently predicted the extent to which the secret was judged as harmful to well-being (Studies 2a-c and 3). Reciprocally, reminding participants of the ways in which a secret does not need to be harmful (i.e., across the three dimensions of secrets) bolstered participants' feelings of well-being and efficacy with regard to coping with that secret (Study 4). A final study that examined secrets from romantic partners replicated the effect on perceived coping efficacy, which in turn predicted daily indicators of relationship quality (Study 5).