# Use select_dtypes with np.number for select all numeric columns: cols = [df.select_dtypes([np.number]).columns] print (list(cols)) # Here is possible specify float64 and int64: cols = [df.select_dtypes([np.int64,np.float64]).columns] print (cols)