diff --git a/src/daft-csv/src/read.rs b/src/daft-csv/src/read.rs
index 68c5ebe144..850b65d711 100644
--- a/src/daft-csv/src/read.rs
+++ b/src/daft-csv/src/read.rs
@@ -289,7 +289,7 @@ async fn read_csv_single_into_table(
             .zip(&read_fields)
             .map(|(array, field)| Series::try_from((field.name.as_ref(), array)))
             .collect::<DaftResult<Vec<Series>>>()?;
-        let table = Table::new(owned_read_schema.clone(), columns_series)?;
+        let table = Table::new_unchecked(owned_read_schema.clone(), columns_series);
         if let Some(predicate) = &predicate {
             let filtered = table.filter(&[predicate.as_ref()])?;
             if let Some(include_columns) = &include_columns {
diff --git a/src/daft-table/src/lib.rs b/src/daft-table/src/lib.rs
index 154e53d991..20366460db 100644
--- a/src/daft-table/src/lib.rs
+++ b/src/daft-table/src/lib.rs
@@ -71,6 +71,14 @@ impl Table {
         })
     }
 
+
+    pub fn new_unchecked<S: Into<SchemaRef>>(schema: S, columns: Vec<Series>) -> Self {
+        Table {
+            schema: schema.into(),
+            columns: columns,
+        }
+    }
+
     pub fn empty(schema: Option<SchemaRef>) -> DaftResult<Self> {
         match schema {
             Some(schema) => {