diff --git a/DESCRIPTION b/DESCRIPTION
index d3f912d8..332a98ec 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -39,6 +39,8 @@ Suggests:
     RWeka,
     stream,
     testthat (>= 3.0.0)
+Remotes:
+    mlr-org/mlr3
 Config/testthat/edition: 3
 Encoding: UTF-8
 Roxygen: list(markdown = TRUE, r6 = TRUE)
diff --git a/NAMESPACE b/NAMESPACE
index a0e4519c..8e8e87e1 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -10,6 +10,7 @@ S3method(as_task_clust,data.frame)
 S3method(as_task_clust,formula)
 S3method(c,PredictionDataClust)
 S3method(check_prediction_data,PredictionDataClust)
+S3method(create_empty_prediction_data,TaskClust)
 S3method(filter_prediction_data,PredictionDataClust)
 S3method(is_missing_prediction_data,PredictionDataClust)
 export(LearnerClust)
diff --git a/R/PredictionDataClust.R b/R/PredictionDataClust.R
index 5047eaba..0b339163 100644
--- a/R/PredictionDataClust.R
+++ b/R/PredictionDataClust.R
@@ -89,3 +89,19 @@ filter_prediction_data.PredictionDataClust = function(pdata, row_ids, ...) {
 
   pdata
 }
+
+#' @export
+create_empty_prediction_data.TaskClust = function(task, learner) {
+  predict_types = mlr_reflections$learner_predict_types[["clust"]][[learner$predict_type]]
+
+  pdata = list(
+    row_ids = integer(),
+    partition = integer()
+  )
+
+  if ("prob" %in% predict_types) {
+    pdata$prob = matrix(integer())
+  }
+
+  set_class(pdata, "PredictionDataClust")
+}
diff --git a/tests/testthat/test_PredictionClust.R b/tests/testthat/test_PredictionClust.R
index b694ab6e..c005987b 100644
--- a/tests/testthat/test_PredictionClust.R
+++ b/tests/testthat/test_PredictionClust.R
@@ -25,3 +25,27 @@ test_that("filter works", {
   expect_set_equal(pdata$row_ids, 1:3)
   expect_integer(pdata$partition, len = 3)
 })
+
+test_that("construction of empty PredictionDataClust", {
+  task = tsk("usarrests")
+
+  learner = lrn("clust.featureless", predict_type = "partition")
+  learner$train(task)
+  pred = learner$predict(task, row_ids = integer())
+  expect_prediction(pred)
+  expect_set_equal(pred$predict_types, "partition")
+  expect_integer(pred$row_ids, len = 0L)
+  expect_numeric(pred$partition, len = 0L)
+  expect_null(pred$prob)
+  expect_data_table(as.data.table(pred), nrows = 0L, ncols = 2L)
+
+  learner = lrn("clust.featureless", predict_type = "prob")
+  learner$train(task)
+  pred = learner$predict(task, row_ids = integer())
+  expect_prediction(pred)
+  expect_set_equal(pred$predict_types, c("partition", "prob"))
+  expect_integer(pred$row_ids, len = 0L)
+  expect_numeric(pred$partition, len = 0L)
+  expect_numeric(pred$prob, len = 0L)
+  expect_data_table(as.data.table(pred), nrows = 0L, ncols = 3L)
+})