This paper addresses the issue of measuring the NAIRU for the euro area and assessing the robustness and precision of the obtained estimates. The empirical framework adopted is based on systems combining an Okun-type relationship between cyclical unemployment and the output gap with a Phillips curve and stochastic laws of motion for the NAIRU and potential output. Such systems have been estimated using Kalman-filter techniques. The aim of this approach is double: on the one hand, recent advances in the mentioned techniques are exploited, also with the aim of assessing the degree of uncertainty around the derived measures; on the other hand, the robustness of the approach is tested by looking at alternative versions of the systems themselves. The results obtained point to an estimate of the area-wide NAIRU that is robust to changes in the underlying models. This robustness is shown to hold both in terms of the mean - i.e., the shape of the resulting NAIRU - and the variance of the process. The latter is derived through bootstrap exercises using the models alone or pooled together. The evidence found suggests that the increase in the aggregate NAIRU that took place in the early part of the sample period has come to a halt and may be about to be reversed. The estimated final level for the NAIRU is around 10 per cent. Furthermore, the bootstrap exercises point to confidence bands close to 1 per cent around the estimated value.