Knut Are Aastveit, Claudia Foroni, and Francesco Ravazzolo, "Density Forecasts with MIDAS Models", Journal of Applied Econometrics, Vol. 32, No. 4, 2017, pp. 783-801. Data files are provided as ASCII files in DOS format (extensions .csv) and as Excel files (extensions .xlsx). The former are zipped in the file afr-data.zip, and the latter are zipped in the file afr-xlsx.data. Unix/Linux users should use "unzip -a" when unzipping afr-data.zip only. The files described below contain the various time series of data used in the article, with rows corresponding to dates and columns to variables. The first column in each file provides the dates of the observation. The first row lists the variables. The files GDP_1-GDP_3 contain the monthly vintages of the quarterly real output data, downloaded from the real-time data set for macroeconomist at the website of the Federal Reserve Bank of Philadelphia. The vintages are from 2001m4 to 2015m9 (in total 174 vintages) with observations from 1983q4 to 2015q2. The monthly vintages refer to the being in the middle of the month. Typically that is before the BEA releases its next report towards the end of the month. That means that a vintage dated in "January" includes the BEA's report released in December. The first row of each file lists the vintage variable names. For example, ROUTPUT01M4 is the vintage for 2001m4. In the article we construct quarterly real output growth as the log difference of the real GDP series. The data vintages 2015m8 and 2015m9 are only included for GDP (and not for the other variables). The reason for this is that we use the second available estimate of GDP as the actual measure for the forecasting evaluation period 2001q3-2015q2,. The second release of GDP for 2015q2 is first available in the 2015m9 vintage. The files EMP_1-EMP_3, IP_1-IP_3 and CU_1-CU_3 contain monthly vintages of monthly nonfarm payroll employment, the total industrial production index and the total capital utilization rate, respectively. These data series and vintages are also downloaded from the real-time data set for macroeconomist at the website of the Federal Reserve Bank of Philadelphia. The vintages are from 2001m4 to 2015m7 (in total 172 vintages) with observations from 1983m12 to 2015M6. The monthly vintages refer to the information after the data release in each month. EMP is typically released on the first Friday of each month, while IP and CU are released around the middle of the month. The first row of each file lists the vintage variable names. For example, EMPLOY01M4 is the vintage for 2001m4. In the article we convert the monthly indicators to monthly growth rates by taking the log difference of the series. The files CFNAI_1-CFNAI_3 contain monthly vintages with monthly observations of the Chicago Fed National Activity Index, downloaded from the website of the Federal Reserve Bank of Chicago. The vintages are from 2001m4 to 2015m7 (in total 172 vintages) with observations from 1983m12 to 2015M6. The monthly vintages refer to the information after the data release in each month. The CFNAI is typically released between the 22-25h of the month. The first row of each file lists the vintage variable names. For example, CF42001 is the vintage for 2001m4. The file NFCI_weekly contains final actual series of weekly observations of the Chicago Fed National Financial Condition Index, downloaded from the website of the Federal Reserve Bank of Chicago. Real time vintages of the NFCI are not available. However, data revisions in the NFCI tend to be small and are more pronounced near the beginning of each month. The third column indicates the number for the weak of the month. In the article we report results using information that includes the third week of the month.