There are some guides and syntax for using multiple imputation in SPSS that can be found here. You can find some information on the SPSS Developer site too.
Update: My take on using multiply imputed values in SPSS: It is way too much of a hassle compared to other programs (e.g. Stata). I spent hours tailoring one of the programs linked above, then was able to have it done in Stata in minutes. The output was more understandable and included more useful information. While people who are very good SPSS programmers might be able to do it reasonably quickly, this was not at all my experience. Bottom line: I have not found an easy way to use multiply imputed values within a complex sample survey in SPSS (my work was with the NS-CSHCN). Please let me know if you find otherwise!