How to read an Excel file with pandas

Reading an Excel workbook with pandas turns a worksheet into a DataFrame without manually exporting the data to CSV first. It fits workbook-based handoffs where Python needs the rows for cleaning, filtering, joining, or reporting.

The pandas.read_excel() function reads the first sheet by default, and sheet_name selects a worksheet by name or zero-based position. For modern .xlsx files, openpyxl is the usual reader engine when engine=None lets pandas choose from the file extension.

Excel imports need explicit choices when a workbook has multiple sheets, helper columns, text-like IDs, or values that should become missing data. A checked import should show the intended worksheet, selected columns, preserved identifiers, parsed dates, expected row count, and dtypes before downstream analysis starts.

1. Install the openpyxl reader package in the Python environment that runs pandas.

   $ python3 -m pip install openpyxl

   pandas uses openpyxl for newer .xlsx workbooks unless another engine is selected. Install pandas too when the environment is new.
   Related: How to install pandas with pip

2. Create a sample workbook generator.

   make_sales_workbook.py

   import pandas as pd
    
   orders = pd.DataFrame(
       {
           "Order ID": ["01001", "01002", "01003"],
           "Customer": ["Ada Lovelace", "Lin Chen", "Maya Patel"],
           "Region": ["EMEA", "APAC", "AMER"],
           "Total": [149.50, "pending", 212.00],
           "Order Date": ["2026-06-01", "2026-06-02", "2026-06-03"],
           "Internal Note": ["reviewed", "awaiting invoice", "reviewed"],
       }
   )
    
   lookup = pd.DataFrame(
       {
           "Region": ["EMEA", "APAC", "AMER"],
           "Owner": ["Finance Europe", "Finance Asia", "Finance Americas"],
       }
   )
    
   with pd.ExcelWriter("sales.xlsx", engine="openpyxl") as writer:
       orders.to_excel(writer, sheet_name="Orders", index=False)
       lookup.to_excel(writer, sheet_name="Lookup", index=False)
    
   print("Created sales.xlsx")
   print("Sheets: Orders, Lookup")

   The sample workbook includes two sheets, one helper column, a text-like order ID column, and one pending value so the reader options have visible effects.

3. Run the workbook generator.

   $ python3 make_sales_workbook.py
   Created sales.xlsx
   Sheets: Orders, Lookup

4. List the workbook sheets before choosing one.

   $ python3 - <<'PY'
   import pandas as pd
   
   with pd.ExcelFile("sales.xlsx") as workbook:
       print(workbook.sheet_names)
   PY
   ['Orders', 'Lookup']

   ExcelFile is useful when a workbook has several sheets or when sheet names need to be checked before importing data.

5. Create the Excel reader script.

   read_sales_excel.py

   import pandas as pd
    
   df = pd.read_excel(
       "sales.xlsx",
       sheet_name="Orders",
       engine="openpyxl",
       usecols=["Order ID", "Customer", "Region", "Total", "Order Date"],
       dtype={
           "Order ID": "string",
           "Customer": "string",
           "Region": "string",
       },
       parse_dates=["Order Date"],
       na_values=["pending"],
   )
    
   df = df.rename(
       columns={
           "Order ID": "order_id",
           "Customer": "customer",
           "Region": "region",
           "Total": "total_usd",
           "Order Date": "order_date",
       }
   )
    
   print(df.to_string(index=False))
   print()
   print(df.dtypes)
   print()
   print(f"Rows: {len(df)}")
   print(f"Columns: {', '.join(df.columns)}")

   usecols selects workbook column labels before the rename. dtype preserves leading zeros in Order ID, parse_dates converts the date column, and na_values treats pending as missing data.

6. Run the reader script.

   $ python3 read_sales_excel.py
   order_id     customer region  total_usd order_date
      01001 Ada Lovelace   EMEA      149.5 2026-06-01
      01002     Lin Chen   APAC        NaN 2026-06-02
      01003   Maya Patel   AMER      212.0 2026-06-03
   
   order_id              string
   customer              string
   region                string
   total_usd            float64
   order_date    datetime64[us]
   dtype: object
   
   Rows: 3
   Columns: order_id, customer, region, total_usd, order_date

   Expect three rows, five selected columns, a preserved leading zero in each order_id value, a missing value for the pending total, and a parsed order_date dtype.

7. Remove the sample files after the import behavior is confirmed.

   $ rm sales.xlsx make_sales_workbook.py read_sales_excel.py