There are many reasons why your system may crash due to a DRIVER_IRQL_NOT_LESS_OR_EQUAL STOP or BSOD (Blue Screen of Death) error, so this guide will provide some common tips that may solve the problem, along with some troubleshooting advice. This error message means that the system attempted to access pageable memory using a kernel interrupt request level (IRQL) that was too high. This is one of the most common errors and has many possible causes, but the most common scenarios are: Faulty drivers Hardware issue Anti-virus software problem Overclocking too far If you notice any filenames mentioned on the blue screen, googling for the filename may indicate if it is attached to any driver or software package. This can pinpoint the driver which you need to update. SCSI, Network and graphics drivers are often the offenders if this is a driver problem - so you can try updating all your drivers manually if it is difficult to track down the problem. It is possible to analyse a "dump file" which contains more detailed information about the crash. These dump file are located in C:\Windows\MiniDump and can be opened using diagnostic packages or posted to the forums for analysis by an expert. This extra information will often provide a good clue to the BSOD crash. If you aren't able to boot in to Windows 7 to fix the problem, you may be able to boot using safe mode (press F8 during startup, or follow this guide) and access Windows again. If this still doesn't work, try selecting the "Last known good configuration" option. If this also fails, using system restore to roll back your system a few days may be a valid option. You can use your Windows 7 installation DVD to boot to the recovery options menu, then select "System Restore". Hardware issues are more difficult to diagnose, but if you have resorted to a clean install of Windows 7 and the problem still occurs, then it is a possibility. Memory is a frequent culprit, so you can use a tool like MemTest (or the Windows 7 memory diagnostics) to check your RAM for problems. Replacing components is often the only foolproof way to check for faults.