A key prediction of standard models of economic growth is that the output-capital ratio is constant along the economy s balanced growth path. Using data for 16 OECD countries over 135 years we examine whether the output-capital ratio reverts to a constant in the long run using univariate and panel stationarity tests with structural breaks. Univariate stationarity tests with one break provide more evidence of trend reversion than mean reversion. However, in stationarity tests with two breaks and with up to five breaks, the results are largely independent of the inclusion of a trend. When we allow for up to five breaks in the intercept we find that for 12 of the 16 countries, the output-capital ratio is mean reverting and when we allow for up to five breaks in the intercept for 11 of the 16 countries, the output-capital ratio is trend reverting. We also find that for the panel as a whole the outputa??capital ratio is mean reverting and trend reverting when we allow for up to five breaks.