I've just been trying to do a fresh install of Tiger onto a Powerbook and here's what happened: Initial install of 10.4 fine. Install update - freezes on reboot. Erase disk, repeat install, same freeze on reboot. Zero out disk hoping to map out any bad blocks. Repeat install procedure - freezes on update as before. Start in verbose mode, see "disk I/O error" before blue screen shows and freezes. Start in single user mode, run fsck -yf - it reports all good. Boot from ASD disk, it runs 51 tests - all fine. Try installing Leopard, install update, seems fine but freezes after opening system preferences. Install Leopard again, then install and run TechTool and perform surface scan - sure enough, it finds bad blocks some of with contain OS files. Throughout all of this, prior to freezing the hard drive make a repetitive chugging noise (not clicking) as if seeking endlessly. I've zeroed the drive again and installed Tiger but with no update - at the moment it's fine with no dodgy hard drive noise. I'll later try the updates. My questions are these: Are bad blocks not recognised by SMART utility and ASD disk? When bad blocks are flagged and mapped does that catalogue not survive a fresh OS install? When they have been recognised, shouldn't the OS just work around them? Is it possible the drive is so damaged that it's making new bad blocks as soon as the existing ones are mapped out? UPDATE: I'm trying this - my logic is that 10.4 installs in the space before the bad blocks, the update goes into the bad blocks (whether they've been marked out or not) hence the problem. What I've done is create a 10Gb blank file in Terminal to hopefully 'contain' the bad block area (it's a 100Gb disk so not too much of a hit) - anything I install now should come after it and as it isn't a file of any use, once it's catalogued it shouldn't have to be read or written to again. Well that's my reasoning - whether it's flawed I'll soon find out!