Walter Nissen (et alii), I'm glad to see that I'm not the only one confused by the (apparently) random nature of some TLE sets' ndot2 values! Unfortunately, I've had no luck making any sense of them. Once I thought I could deduce decent values by examining the Mean Anomaly values in different TLEs, but, alas, at least for many of the satellites I care to observe, their eccentricities are so low that it is difficult to precisely locate the perigee position. Since what one observes, at least for a near-circular orbit, is the argument of latitude, the sum of the argument of the perigee and the true anomaly, small, compensating, errors in both of these have almost no effect on predictions. Which to me means that for low-eccentricity orbits, even the MA itself is suspect! I then tried a simple least-squares fit of the perigee position versus epoch, and of the mean argument of latitude versus epoch, but neither approach produced meaningful results! (My subject was Mir and I tried to include element sets over time spans which excluded any [obvious] orbit-tweaking maneuvers.) Keep smiling and observing! Ed Light