Luc Bauwens, Gary Koop, Dimitris Korobilis, and Jeroen Rombouts, "The Contribution of Structural Break Models to Forecasting Macroeconomic Series", Journal of Applied Econometrics, Vol. 30, No. 4, 2015, pp. 596-620. We have 39 original series in total: 18 quarterly (sample is 1959Q1--2011Q3) and 21 monthly series (sample is 1959M1--2011M9). These original, untransformed series can be found in the Excel file DATA_BKKR_rev.xlsx and the csv files DATA_BKKR_rev-m.csv and DATA_BKKR_rev-q.csv. In our forecasting exercises, we also use quarterly versions of our 21 monthly variables by taking the average over each quarter (therefore, we forecast 39 + 21 = 60 time series in total). The files monthly.dat, quarterly.dat and monthly2quarterly.dat contain all the series we use in the forecasting exercises after we apply stationarity transformations. The exact stationarity transformations we applied to each series (mainly first log differences) can be found in Table 2 of the paper. The five .csv and .dat files are ASCII files in DOS format. They are zipped in the file bkkr-txt.zip. Unix/Linux users should use "unzip -a". The file DATA_BKKR_rev.xlsx is zipped in the file bkkr-xlsx.zip. All series used in this paper were downloaded from the Federal Reserve Economic Data (FRED) website in November 2011, and they are freely available at http://research.stlouisfed.org/fred2/. For further questions about the data, please contact Dimitris Korobilis (Dimitris.Korobilis [AT] glasgow.ac.uk) or Jeroen Rombouts (b00469813 [AT] essec.edu). Luc Bauwens Gary Koop Dimitris Korobilis Jeroen Rombouts