我们采用的是这样一种方式：

for(int i = table.Rows.Count - 1; i >= 0; i--) {
    DataRow row = table.Rows[i];
    if ( row.RowState == DataRowState.Deleted ) { table.Rows.RemoveAt(i); }
}

这将满足任何FK级联关系，如'delete'（即DataTable.Clear（）不会）：

DataTable dt = ...;
// Remove all
while(dt.Count > 0)
{
     dt.Rows[0].Delete();
}

dt.Rows.Clear();
dt.Columns.Clear();   //warning: All Columns delete
dt.Dispose();

我通常执行以下SQL命令：

DELETE FROM TABLE WHERE ID>0

因为您使用的是SQL Server数据库，所以我建议您只执行SQL命令"DELETE FROM " + dt.TableName。

我会删除该表，这是删除所有内容的最快方法。然后重新创建该表。

您可以在SQL Server数据库上创建一个存储过程来删除表中的所有行，从C#代码中执行该过程，然后重新查询数据表。

Here is the solution that I settled on in my own code after searching for this question, taking inspiration from Jorge's answer.

DataTable RemoveRowsTable = ...;
int i=0;
//Remove All
while (i < RemoveRowsTable.Rows.Count)
{
     DataRow currentRow = RemoveRowsTable.Rows[i];
     if (currentRow.RowState != DataRowState.Deleted)
     {
         currentRow.Delete();
     }
     else
     {
         i++;
     }
}

This way, you ensure all rows either get deleted, or have their DataRowState set to Deleted.
Also, you won't get the InvalidOperationException due to modifying a collection while enumerating, because foreach isn't used. However, the infinite loop bug that Jorge's solution is vulnerable to isn't a problem here because the code will increment past a DataRow whose DataRowState has already been set to Deleted.