You can access the index attribute of a df using df.index[i]
>> import pandas as pd
>> import numpy as np
>> df = pd.DataFrame({'a':np.arange(5), 'b':np.random.randn(5)})
a b
0 0 1.088998
1 1 -1.381735
2 2 0.035058
3 3 -2.273023
4 4 1.345342
>> df.index[1] ## Second index
>> df.index[-1] ## Last index
>> for i in xrange(len(df)):print df.index[i] ## Using loop
...
0
1
2
3
4