The hypothesis that the risk of childhood leukemia can be associated with exposure to specific combinations of power frequency and static magnetic fields (SMFs) was evaluated. The data used for this analysis was obtained in a previously reported case control study of childhood cancer in Los Angeles County, California. Time weighted average (TWA) extremely low frequency (ELF) magnetic field and SMF exposures were measured in the bedrooms of 223 subjects: 124 cases with leukemia and 99 controls. The Wertheimer-Leeper (WL) code for wiring configurations was also analyzed as a surrogate measure of 60 Hertz (Hz) ELF magnetic field exposure: WL code data were available for 132 cases and 101 controls. Analysis of laboratory data indicated that 19 bands of SMFs, each having a bandwidth of 9.1 microTesla (microT) together with the 60Hz magnetic field would, be expected to have biological activity. Among the subjects in the case control study, 26 cases and 20 controls were exposed to SMFs lying in bands centered at 38.0 and 50.6microT that were predicted to have biological activity. An increasing trend of leukemia risk was found for subjects whose exposures for at least 80% of the 24 hour monitoring period fell within the 38.0 and 50.6microT bands when assessed by the usual case control techniques. The odds ratio (OR) and its 95% confidence interval (CI) for subjects whose exposures lay within these bands was 3.3 (0.4-30.5 CI) for those subjects exposed to the highest ELF levels compared to subjects exposed to SMFs and 60Hz fields that lay outside these bands. No association between leukemia risk and ELF magnetic fields was observed for subjects whose SMF exposures lay outside the 38.0 and 50.6microT bands; ORs were consistently below 1. When the magnetic field exposures were assessed by the WL wiring codes, the trend in leukemia risk within the 38.0 and 50.6microT bands was statistically significant for the very high current configuration, OR 9.2 (1.3- 64.6 CI; 10 cases, one control). The authors conclude that although the risk estimates were based on limited magnetic field measurements for a small number of subjects, the analysis supports the hypothesis that the risk of childhood leukemia may be related to the combined effects of static and ELF magnetic field exposure.