The US government needs to improve its procedure for generating the list of its satellites that it registers with the UN. This will be a long, ranty, thread. Since the start of the space age it has been standard practice that all Earth satellite launches, even `secret' ones, have been publicly announced. In 1962, around the time of the Cuban Missile Crisis, the US was alarmed when it discovered the USSR had launched several unannounced satellites into low orbit. Were they, perhaps, orbital nukes? No, it turned out, they were failed interplanetary probes. Nevertheless this reinforced the idea that unannounced launches were a bad idea. Relatedly, a UN register of space objects was established; UN Resolution 1721B and, later, the UN Registration Convention came into force. The US has mostly, but not perfectly, complied with the requirements of the convention. Recently there's been a new problem in the US registrations. The registration documents are clearly generated from the US satellite catalog maintained by SpaceForce 18SPCS. This is not good enough. Why? Let's look at an example - the Electron 4 mission of Dec 2018, launch 2018-104. This placed 11 United States payloads in orbit, and two New Zealand rocket stages. All 13 objects are being tracked by Space Force, and we know which 2 objects are the rocket stages. So, we have 11 tracked objects that we know are the 11 US payloads. But, for 5 of these payloads we don't know which object is which. Tracked objects E, F, K, L and P are the satellites Ceres, RSat, Cubesail, Albus and NMTSat in some order, but we don't know which order. In the US satellite catalog, catalog numbers corresponds to objects being tracked, and they only get a name associated with them if Space Force knows what the tracked object is. So, if you search for the NASA 'Albus' satellite or the New Mexico Tech 'NMTSat' in the satellite catalog you won't find anything. If you count the number of US payloads launched in 2018, these won't be included and you'll get too low an answer. More recently, there are the V-R3X payloads from launch 2021-006 which aren't even identified as US objects in the catalog, even though we know they are objects BC, BG and DQ in some order. So, they are not included in US registration document ST/SG/SER.E/977 - but they should be. This is (arguably) not Space Force's fault. It's their catalog and this is what works for them, I guess. BUT: the requirement on the State Dept. for UN registration is NOT 'tell me which of your satellites you are tracking'. It's 'tell me which satellites you have put in orbit'. So it is NOT OK for them to use the satcat as a complete list for this purpose - because it is not OK for them to leave out satellites like Albus just because we don't know which catalog number it is. We know the US launched Albus and it is in orbit, so it should be registered with the UN. I would argue the same is true of the Thinsats deployed from Cygnus even though SpaceForce decided they weren't going to bother cataloging them at all because they reentered after a few days. They met the usual criterion for registration with the UN and should be registered. This used not to be a big problem because very few launched US payloads were not identified with a catalog number. But now with all these cluster cubesat launches, it's much more common. I estimate there are 28 unregistered US satellites in 2018, and 9 in 2019, although only 1 in 2020. There look to be a bunch in 2021 but not all the reg documents are in yet. One possibility is that Space-Track could maintain a separate public table of object information versus launch number for objects known to be in orbit but not associated with a catalog number. It shouldn't be too hard to do better. I keep a list of the unregistered satellites on my website (objects marked [US] in square parentheses in https://planet4589.org/space/gcat/data/cat/psatcat.html If one person can do this in his spare time, it seems like the US government should be able to manage it. Do better, America! End of rant.