diff --git a/samples/Scala Connector with Service Principal using MSAL.scala b/samples/Scala Connector with Service Principal using MSAL.scala
new file mode 100644
index 0000000..1a1a649
--- /dev/null
+++ b/samples/Scala Connector with Service Principal using MSAL.scala	
@@ -0,0 +1,47 @@
+import com.microsoft.aad.msal4j.ClientCredentialFactory
+import com.microsoft.aad.msal4j.ClientCredentialParameters
+import com.microsoft.aad.msal4j.ConfidentialClientApplication
+import com.microsoft.aad.msal4j.IAuthenticationResult
+
+import com.nimbusds.oauth2.sdk.http.HTTPResponse
+import com.nimbusds.oauth2.sdk.id.ClientID
+
+import java.util.Collections
+
+
+val dbname = "yourdbname"
+val servername = "jdbc:sqlserver://yourazsqlserver.database.windows.net"
+val tablename = "yourtablename"
+
+
+val authority="https://login.microsoftonline.com/yourtenantid"
+val clientId="clientidofyourserviceprincipal"
+val secret="clientsecret"
+val scope="https://database.windows.net/.default"
+
+val app = ConfidentialClientApplication.builder(
+                clientId,
+                ClientCredentialFactory.createFromSecret(secret))
+                .authority(authority)
+                .build()
+// With client credentials flows the scope is ALWAYS of the shape "resource/.default", as the
+// application permissions need to be set statically (in the portal), and then granted by a tenant administrator
+val clientCredentialParam = ClientCredentialParameters.builder(
+                Collections.singleton(scope))
+                .build()
+
+val accessToken=app.acquireToken(clientCredentialParam).get().accessToken()
+
+val df = spark.read.format("com.microsoft.sqlserver.jdbc.spark")
+  .option("url",servername)
+  .option("dbtable",tablename)
+  .option("databaseName",dbname)
+  .option("accessToken",accessToken)
+  .option("connectTimeout",30) //seconds, allow extra time for paused DB to start-up
+  .option("queryTimeout",30)  //seconds
+  .option("hostNameInCertificate","*.database.windows.net")
+  .option("trustServerCertificate","true")
+  .option("encrypt","true")
+  .load()
+display(df)
+