Premarital education may help emerging adults form healthy relationships, but evaluation research is needed, particularly with community samples. We studied emerging adults in the Premarital Interpersonal Choices and Knowledge (PICK) program, using a pre‐ to post‐ and a posttest‐then‐retrospective‐pretest design to examine change in perceived relationship skills, partner selection, relational patterns, and relationship behaviors and attitudes. Mixed models analyses showed that scores for the treatment group (n = 682) increased from pre to post on all four outcomes. Changes in scores for the nonequivalent comparison group (n = 462) were nonsignificant. In addition, significant differences between pre‐ and retrospective prescores demonstrated evidence for response shift bias. The results suggest that the PICK program helps participants increase their knowledge regarding the components of healthy relationship formation.