diff --git a/inst/csv/replacementPatterns.csv b/inst/csv/replacementPatterns.csv
index ab29fed..8a14f1c 100644
--- a/inst/csv/replacementPatterns.csv
+++ b/inst/csv/replacementPatterns.csv
@@ -1140,6 +1140,7 @@ spark,"WITH @cte AS (SELECT @s1 '@literal' @s2)","WITH @cte AS (SELECT @s1 CAST(
 spark,UPDATE STATISTICS @a;,
 spark,"SELECT @columns FROM (@a) @x,(@b) @y;","SELECT @columns FROM (@a) @x cross join (@b) @y;"
 spark,"ALTER TABLE @table ADD COLUMN @([\w_-]+)column @type DEFAULT @default;","ALTER TABLE @table ADD COLUMN @column @type; \nALTER TABLE @table SET TBLPROPERTIES('delta.feature.allowColumnDefaults' = 'supported'); \nALTER TABLE @table ALTER COLUMN @column SET DEFAULT @default;"
+spark,"CAST(@a AS DATE)","IF(try_cast(@a AS DATE) IS NULL, to_date(cast(@a AS STRING), 'yyyyMMdd'), try_cast(@a AS DATE))"
 sqlite extended,"IIF(@condition, @whentrue, @whenfalse)","CASE WHEN @condition THEN @whentrue ELSE @whenfalse END"
 sqlite extended,"ROUND(@a,@b)","ROUND(CAST(@a AS REAL),@b)"
 sqlite extended,+ '@a',|| '@a'
diff --git a/tests/testthat/test-translate-spark.R b/tests/testthat/test-translate-spark.R
index d38d5f0..ef180b9 100644
--- a/tests/testthat/test-translate-spark.R
+++ b/tests/testthat/test-translate-spark.R
@@ -448,3 +448,8 @@ test_that("translate sql server -> spark add column with default", {
   sql <- translate("ALTER TABLE mytable ADD COLUMN mycol int DEFAULT 0;", targetDialect = "spark")
   expect_equal_ignore_spaces(sql, "ALTER TABLE mytable ADD COLUMN mycol int; \n ALTER TABLE mytable SET TBLPROPERTIES('delta.feature.allowColumnDefaults' = 'supported'); \n ALTER TABLE mytable ALTER COLUMN mycol SET DEFAULT 0;")
 })
+
+test_that("translate sql server -> spark cast string as date", {
+  sql <- translate("SELECT CAST('20191201' AS DATE);", targetDialect = "spark")
+  expect_equal_ignore_spaces(sql, "SELECT IF(try_cast('20191201' AS DATE) IS NULL, to_date(cast('20191201' AS STRING), 'yyyyMMdd'), try_cast('20191201' AS DATE));")
+})