There are few (if any) population-based prospective studies that provide information on factors associated with temporal sequencing of suicide. In this prospective population-based study, the National Health Interview Survey (NHIS), 1986-1994, was linked to the National Death Index (NDI), 1986-1997, to assess factors that predict recent (within 12 months of interview, termed "sooner") suicide versus suicide further in the future (more than 12 months after interview, termed "later"). Of the 653 completed suicides in the NHIS cohort, 13.4 percent completed suicide within a year of interview, and 86.6 percent did so after a year. Sooner decedents were more likely to be White, less educated, unemployed, and to use firearms than any other method compared with later decedents. Surprisingly, sooner decedents had higher levels of self-rated health at baseline. These results have substantial implications for clinicians and other professionals who interact with people at highest risk of suicide. Unfortunately, it may be unrealistic to expect that health care providers can modify the behavior of individuals at highest risk of suicide.