Objective Attachment research has traditionally focused on traits, enhancing our understanding of attachment-related individual differences. However, to chart the dynamic properties of the attachment system, more research is needed on the within-person fluctuation of attachment states. In this ecological momentary assessment (EMA) study, we examined (a) the associations between the baseline, variability, and inertia of each state attachment dimension (security, avoidance, and anxiety) and (b) how trait attachment (anxiety and avoidance) predicts these dynamic features.

Method In two adult samples (Ns = 122 and 127), trait attachment dimensions were first assessed using Experiences in Close Relationships-Revised. Then, attachment states were assessed seven or ten times a day over 1 week (4629 and 5322 successful EMA observations).

Results For state security, individuals with high baseline exhibited lower variability. In contrast, for state avoidance, those with high baseline showed higher variability. Both trait attachment anxiety and avoidance predicted lower baseline and higher variability of state security. Moreover, both trait dimensions predicted higher baselines of the corresponding states.

Conclusions Our findings provide insights into the real-time regulatory dynamics of the attachment system and their interconnection with trait attachment, underscoring the importance of baseline and variability in understanding how attachment manifests in everyday life.