CRC fuckers.

Could also be a problem with the IDE interface...

Jam it back in, in the dark.

Also check SMART data with something like SpeedFan. Especially the 'reallocated sector count' is important. If you're drive is failing the probably the SMART diagnosis will tell you.
Best thing at this time would be to transfer data to another drive.

There's nowhere I can't reach.

No, as soon as the reallocated sector count (in RAW) is different from zero (or the value the drive had when you purchased it - sometimes the hd vendor encodes some more data into this special field) the drive is beginning to fail.

SMART data from my internal Fujitsu harddisk:

ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   100   100   046    Pre-fail  Always       -       249561
  2 Throughput_Performance  0x0004   100   100   000    Old_age   Offline      -       11862016
  3 Spin_Up_Time            0x0003   100   100   025    Pre-fail  Always       -       1
  4 Start_Stop_Count        0x0032   098   098   000    Old_age   Always       -       5749
  5 Reallocated_Sector_Ct   0x0033   100   100   024    Pre-fail  Always       -       8589934592000
  7 Seek_Error_Rate         0x000e   100   100   000    Old_age   Always       -       3111
  8 Seek_Time_Performance   0x0004   100   100   000    Old_age   Offline      -       0
  9 Power_On_Hours          0x0032   083   083   000    Old_age   Always       -       31778553
 10 Spin_Retry_Count        0x0012   100   100   000    Old_age   Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   000    Old_age   Always       -       684
192 Power-Off_Retract_Count 0x0032   100   100   000    Old_age   Always       -       19
193 Load_Cycle_Count        0x0032   097   097   000    Old_age   Always       -       62088
194 Temperature_Celsius     0x0022   100   100   000    Old_age   Always       -       35 (Lifetime Min/Max 10/51)
195 Hardware_ECC_Recovered  0x001a   100   100   000    Old_age   Always       -       240
196 Reallocated_Event_Count 0x0032   100   100   000    Old_age   Always       -       454098944
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age   Always       -       0
200 Multi_Zone_Error_Rate   0x000e   100   100   000    Old_age   Always       -       14384
203 Run_Out_Cancel          0x0002   100   100   000    Old_age   Always       -       433771707999

Important are the reallocated sector count (mind the RAW value), the current pending sector count, reallocated event count (and everything else that has to do with reallocation), the spin up time (if this increases then you're soon going to have some mechanical damage), the CRC error counts (indicates failing electronics) and of course the temperature (higher temperature also indicates mechanical damage producing more heat than normal).

Why are the reallocation values so important. The drives transparently replace defective (better to say: nearly defective) with some spare sectors (there is a limited number normally not available to the user).
Now experience shows that if reallocation occurs (sectors are becoming defective) it doesn't take long for the next sector to become defective. Reasons are that the material of the platter has almost homogenous quality, so IF some sectors fails (because of the age of the material) then all other sectors (which lie on material has nearly the same age and therefore same quality) are also likely to fail. This does not have to be the case but it's safer then saying "hey it's only one lousy sector" and a week later there is already a count of around 1000 (the amount of spare sectors is limited!).
Another reason why one should act quick is that if the sector reallocation was triggered by some (non-drive-mechanic-critical) headcrash. The crash destroys the information of the sector (or sectors...) but furthermore can produce tiny particles (scratched-off material) that are now inside the drive casing...

This thing is sticky, and I don't like it. I don't appreciate it.