OBJECTIVE: The aim of this study was to investigate the significance of the combined use of BMI and AEC in reducing the radiation dose of CT volume scans of the lumbar spine. METHODS: A prospective study was performed to continuously collect data from 50 patients (age range from 19 to 60 years, male versus female 20/30) whose BMIs were less than 25 kg/m^{2} (group A) and 50 patients (age range from 21 to 82 years, male versus female 24/26) whose BMIs were equal to or more than 25 kg/m^{2} (group B). The 50 patients in each group were randomly divided into 5 subgroups with each subgroup having lower radiation dose from subgroup 1 to 5. All the patients were performed lumbar spiral CT scans (GE LightSpeed VCT 64-slice scanner) and the scan parameters were different in different subgroups. Volume CT dose index (CTDIvol) was recorded. The qualities of the images were graded. The one-way ANOVA and Kruskal-Wallis test were done. RESULTS: Both in group A and B, there were significant differences in CTDIvol among the 5 subgroups (P< 0.001). The quality of the images in the 5 subgroups of group A didn't show statistical difference. The standard deviation (SD) and signal to noise ratio (SNR) values of the L4-5 psoas major muscles in subgroup 5 of group B was statistical different from the other 4 subgroups (P< 0.01). CONCLUSION: Use of BMI combined with AEC reduces radiation dosage, without compromising the image quality. For patients in group A and group B, parameters of subgroup 5 and subgroup 4 may respectively be applied for lower dose CT scanning. [ABSTRACT FROM AUTHOR]