diff --git a/execution/execmodule/forkchoice.go b/execution/execmodule/forkchoice.go
index 7186fa73322..1b6f67ca84e 100644
--- a/execution/execmodule/forkchoice.go
+++ b/execution/execmodule/forkchoice.go
@@ -290,6 +290,14 @@ func (e *ExecModule) unwindIfNeeded(
 			err = fmt.Errorf("updateForkChoice: %w", err)
 			return nil, err
 		}
+	} else {
+		execProgress, err := stages.GetStageProgress(tx, stages.Execution)
+		if err != nil {
+			return nil, fmt.Errorf("updateForkChoice: %w", err)
+		}
+		if execProgress > unwindTarget {
+			e.logger.Info("updateForkChoice: unwind skipped with executed state above reorg point", "unwindTarget", unwindTarget, "lastCanonicalBlock", lastCanonicalBlock, "execProgress", execProgress)
+		}
 	}
 	// SD.Unwind (inside RunUnwind) tx-aware-invalidates the BranchCache by
 	// the unwound txNum, so no whole-cache clear is needed here.
diff --git a/execution/stagedsync/stage_execute.go b/execution/stagedsync/stage_execute.go
index 4fa4990f57f..0a7bdc94e8d 100644
--- a/execution/stagedsync/stage_execute.go
+++ b/execution/stagedsync/stage_execute.go
@@ -407,6 +407,7 @@ func UnwindExecutionStage(u *UnwindState, s *StageState, doms *execctx.SharedDom
 		// Do not `ResetPendingUpdates()` here. Unlike the disk-unwind path below (which discards then
 		// rebuilds commitment state via unwindExec3 + SeekCommitment), this early return only rewinds the in-RAM overlay
 
+		logger.Info(fmt.Sprintf("[%s] Unwind Execution: RAM-only, disk state untouched", u.LogPrefix()), "unwindPoint", u.UnwindPoint, "progress", s.BlockNumber)
 		_, err = unwindDomsToBlock(ctx, rwTx, cfg.blockReader, doms, s.BlockNumber, nil)
 		return err
 	}
diff --git a/execution/stagedsync/sync.go b/execution/stagedsync/sync.go
index 640d62d1174..91a92181028 100644
--- a/execution/stagedsync/sync.go
+++ b/execution/stagedsync/sync.go
@@ -140,6 +140,7 @@ func (s *Sync) UnwindTo(unwindPoint uint64, reason UnwindReason, tx kv.Tx) error
 			// Ignore in the case that snapshots are ahead of commitment, it will be resolved later.
 			// This can be a problem if snapshots include a wrong chain so it is ok to ignore it.
 			if errors.Is(err, commitmentdb.ErrBehindCommitment) {
+				s.logger.Info("UnwindTo: unwind request dropped, target behind commitment", "requested", unwindPoint, "err", err)
 				return nil
 			}
 			if err != nil {
@@ -538,6 +539,11 @@ func (s *Sync) unwindStage(initialCycle bool, stage *Stage, sd *execctx.SharedDo
 	unwind.Reason = s.unwindReason
 
 	if stageState.BlockNumber <= unwind.UnwindPoint {
+		if stageState.BlockNumber == unwind.UnwindPoint {
+			s.logger.Info("unwind skipped, stage exactly at unwind point", "stage", stage.ID, "unwindPoint", unwind.UnwindPoint)
+		} else {
+			s.logger.Debug("unwind skipped, stage below unwind point", "stage", stage.ID, "progress", stageState.BlockNumber, "unwindPoint", unwind.UnwindPoint)
+		}
 		return nil
 	}