Electrohysterography　has　been　shown　to　provide　relevant　information　on　preventing　preterm　labor.　Recent　studies　have　confirmed　the　feasibility　of　using　the　vectormyometriogram　(VMG)　to　assess　uterine　myoelectric　vector　displacement,　with　different　physiological　implications　for　the　slow　and　fast　waves,　without　suggesting　its　implementation　in　clinical　practice.　The　fast　wave　VMG　component　has　dynamic　behavior　in　any　specific　direction　on　the　X　-Y　plane.　Since　recurrence　is　a　common　feature　of　dynamic　systems,　we　aimed　to　determine　the　recurrence　pattern　of　uterine　vector　displacement,　exploring　its　clinical　potential　in　detecting　imminent　and　preterm　labor　in　women　with　threatened　preterm　labor　and　a　serious　preterm　birth　risk.　For　this,　we　analyzed　the　recurrence　patterns　from　a　2D-vectormyometriogram　using　four　common　statistics:　determinism,　longest　diagonal,　entropy,　and　laminarity.　We　found　significantly　increased　determinism　(0.035　+/-　0.011　vs.　0.077　+/-　0.041),　entropy　(1.768　+/-　0.116　vs.　2.197　+/-　0.24)　and　laminarity　(0.086　+/-　0.034　vs.　0.173　+/-　0.078)　from　the　early　(26-30　weeks)　to　late　(35-37　weeks)　gestation　stages.　As　pregnancy　progresses,　the　uterine　vector　displacement　becomes　more　periodic,　predictable　and　stable,　while　VMG　recurrence　statistics　in　the　fast　wave　high　bandwidth　better　detect　imminent　and　preterm　labor,　outperforming　classical　EHG　parameters　from　bipolar　channels.　The　proposed　method　was　also　resistant　to　motion　artifacts　and　preserved　its　discriminative　capacity　between　the　groups.　Our　results　on　VMG　recurrence　statistics　could　thus　be　another　reliable　biomarker　for　preventing　preterm　labor　in　women　with　threatened　preterm　labor　and　would　favor　transferring　the　EHG　technique　to　clinical　practice.