diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchema.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchema.scala
index 4dd41042856d..afadcf664a0d 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchema.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchema.scala
@@ -22,6 +22,7 @@ import scala.util.control.Exception.allCatch
 import org.apache.spark.rdd.RDD
 import org.apache.spark.sql.catalyst.analysis.TypeCoercion
 import org.apache.spark.sql.catalyst.expressions.ExprUtils
+import org.apache.spark.sql.catalyst.util.DateFormatter
 import org.apache.spark.sql.catalyst.util.TimestampFormatter
 import org.apache.spark.sql.types._
 
@@ -32,6 +33,10 @@ class CSVInferSchema(val options: CSVOptions) extends Serializable {
     options.timeZone,
     options.locale)
 
+  private val dateParser = DateFormatter(
+    options.dateFormat,
+    options.locale)
+
   private val decimalParser = {
     ExprUtils.getDecimalParser(options.locale)
   }
@@ -102,6 +107,7 @@ class CSVInferSchema(val options: CSVOptions) extends Serializable {
           // DecimalTypes have different precisions and scales, so we try to find the common type.
           compatibleType(typeSoFar, tryParseDecimal(field)).getOrElse(StringType)
         case DoubleType => tryParseDouble(field)
+        case DateType => tryParseDate(field)
         case TimestampType => tryParseTimestamp(field)
         case BooleanType => tryParseBoolean(field)
         case StringType => StringType
@@ -152,6 +158,15 @@ class CSVInferSchema(val options: CSVOptions) extends Serializable {
   private def tryParseDouble(field: String): DataType = {
     if ((allCatch opt field.toDouble).isDefined || isInfOrNan(field)) {
       DoubleType
+    } else {
+      tryParseDate(field)
+    }
+  }
+
+  private def tryParseDate(field: String): DataType = {
+    // This case infers a custom `dataFormat` is set.
+    if ((allCatch opt dateParser.parse(field)).isDefined) {
+      DateType
     } else {
       tryParseTimestamp(field)
     }
diff --git a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchemaSuite.scala b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchemaSuite.scala
index c2b525ad1a9f..ac5447b65be1 100644
--- a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchemaSuite.scala
+++ b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchemaSuite.scala
@@ -27,7 +27,8 @@ import org.apache.spark.sql.types._
 class CSVInferSchemaSuite extends SparkFunSuite with SQLHelper {
 
   test("String fields types are inferred correctly from null types") {
-    val options = new CSVOptions(Map("timestampFormat" -> "yyyy-MM-dd HH:mm:ss"), false, "GMT")
+    val options = new CSVOptions(
+      Map("timestampFormat" -> "yyyy-MM-dd HH:mm:ss", "dateFormat" -> "yyyy-MM-dd"), false, "GMT")
     val inferSchema = new CSVInferSchema(options)
 
     assert(inferSchema.inferField(NullType, "") == NullType)
@@ -36,6 +37,8 @@ class CSVInferSchemaSuite extends SparkFunSuite with SQLHelper {
     assert(inferSchema.inferField(NullType, "60") == IntegerType)
     assert(inferSchema.inferField(NullType, "3.5") == DoubleType)
     assert(inferSchema.inferField(NullType, "test") == StringType)
+    // [SPARK-25517] added assert statement for auto inferring DateType
+    assert(inferSchema.inferField(NullType, "2019-03-04") == DateType)
     assert(inferSchema.inferField(NullType, "2015-08-20 15:57:00") == TimestampType)
     assert(inferSchema.inferField(NullType, "True") == BooleanType)
     assert(inferSchema.inferField(NullType, "FAlSE") == BooleanType)
@@ -47,7 +50,8 @@ class CSVInferSchemaSuite extends SparkFunSuite with SQLHelper {
   }
 
   test("String fields types are inferred correctly from other types") {
-    val options = new CSVOptions(Map("timestampFormat" -> "yyyy-MM-dd HH:mm:ss"), false, "GMT")
+    val options = new CSVOptions(
+      Map("timestampFormat" -> "yyyy-MM-dd HH:mm:ss", "dateFormat" -> "yyyy/MM/dd"), false, "GMT")
     val inferSchema = new CSVInferSchema(options)
 
     assert(inferSchema.inferField(LongType, "1.0") == DoubleType)
@@ -55,6 +59,8 @@ class CSVInferSchemaSuite extends SparkFunSuite with SQLHelper {
     assert(inferSchema.inferField(IntegerType, "1.0") == DoubleType)
     assert(inferSchema.inferField(DoubleType, null) == DoubleType)
     assert(inferSchema.inferField(DoubleType, "test") == StringType)
+    // [SPARK-25517] added assert statement for auto inferring DateType
+    assert(inferSchema.inferField(NullType, "2019/03/04") == DateType)
     assert(inferSchema.inferField(LongType, "2015-08-20 14:57:00") == TimestampType)
     assert(inferSchema.inferField(DoubleType, "2015-08-20 15:57:00") == TimestampType)
     assert(inferSchema.inferField(LongType, "True") == BooleanType)
@@ -123,6 +129,8 @@ class CSVInferSchemaSuite extends SparkFunSuite with SQLHelper {
 
     assert(inferSchema.inferField(IntegerType, "\\N") == IntegerType)
     assert(inferSchema.inferField(DoubleType, "\\N") == DoubleType)
+    // [SPARK-25517] added assert statement for auto inferring DateType
+    assert(inferSchema.inferField(DateType, "\\N") == DateType)
     assert(inferSchema.inferField(TimestampType, "\\N") == TimestampType)
     assert(inferSchema.inferField(BooleanType, "\\N") == BooleanType)
     assert(inferSchema.inferField(DecimalType(1, 1), "\\N") == DecimalType(1, 1))