we know about TIRA-like performances.We have set up a realistic simulation scenario for a re-entry campaign. Re-entry predictions and ballistic coefficient calibrations are performed and compared for TIRA-only, EISCAT-only, and both radar available situations. The results are compared in terms of differences in the orbital states over the total observation time span, in the ballistic coefficient estimation, and in the corresponding re-entry epoch.The main conclusion is that, provided a minimum amount of necessary observational information, EISCAT-based re-entry predictions are of comparable accuracy to TIRA-based (but also to GPS-based, and TLE-based if TLE errors are properly accounted) corresponding ones. Even if the worse tracking capabilities of the EISCAT sensor are not able to determine an orbit at the same level of accuracy of TIRA, it turned out that the estimated orbits are anyway equivalent in terms of re-entry predictions, if we consider the relevant parameters involved and their effects on the re-entry time. What happens to be very important is the difficulty in predicting both atmospheric and attitude significant variations in between the current epoch of observation and the actual re-entry. For this reason, it is not easy to keep the actual accuracy of the predictions much lower than 10% of the residual lifetime, apart from particular cases with constant area to mass ratio, and low atmospheric environment variations with respect to current models.A corresponding GOCE ballistic coefficient estimation based on TLE only is presented, and it turns out to be very effective as well. Some critical cases which consist in a minimum amount of observational information, or in difficulties in obtaining OD convergence, are presented, and a list of possible countermeasures is proposed.An experiment with real EISCAT,TIRA, and TLE data is also presented, for the case of 2012-006K AVUM R/B. Additional experiments with simulated trajectories are presented as well, with analogous results.