diff --git a/NEWS.md b/NEWS.md
index 618aecf3..ecb3f7b3 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -5,6 +5,8 @@
 - Support excluding samples or sample groups from defining features with
   *PeakDensity* correspondence analysis (issue #742).
 - Add `plotPrecursorIons()` function.
+- Fix in `dropFeatureDefinitions()` that was not correctly removing additional
+  metadata from gap-filled chromatographic peaks.
 
 
 # xcms 4.1
diff --git a/R/XcmsExperiment.R b/R/XcmsExperiment.R
index 17f4b834..356edbce 100644
--- a/R/XcmsExperiment.R
+++ b/R/XcmsExperiment.R
@@ -1556,9 +1556,8 @@ setMethod(
             object@processHistory, type = .PROCSTEP.PEAK.GROUPING, num = 1L)
         object@featureDefinitions <- .empty_feature_definitions()
         if (.hasFilledPeaks(object)) {
-            object@chromPeaks <- object@chromPeaks[
-                                            !object@chromPeakData$is_filled, ,
-                                            drop = FALSE]
+            object <- .filter_chrom_peaks(
+                object, which(!.chromPeakData(object)$is_filled))
             object@processHistory <- dropProcessHistoriesList(
                 object@processHistory, type = .PROCSTEP.PEAK.FILLING)
         }