A growing body of scholarship in the social sciences shows how household surveys – researchers' main source of income and poverty estimates – are affected by nonresponse and underreporting. Linking Wave 1 of the 2014 Survey of Income and Program Participation to Social Security and IRS administrative records, this paper examines respondents aged 65 and over and the accuracy of their survey responses for earnings, interest, dividends, social insurance, and retirement income. The analysis shows that retirement income – income from pensions and individual retirement accounts – is considerably underreported in SIPP. Substituting administrative records of retirement income increases household income 8.7%, 21.5%, and 27.2% at the 10th, 50th, and 90th percentiles, respectively. SIPP generally captures other forms of income well. The results show that income inequality among those aged 65 and over is higher than SIPP estimates suggest.