Objective To determine the equivalence of SF-12v2 and SF-36v2 for assessing health related quality of life in a general population in Chengdu. Methods The equivalence between SF-12v2 and SF-36v2 was assessed using reliability, validity and responsiveness. Results The eight sub-scales of SF-36v2 had a score ranging from 64.13 to 89.15, compared with a range between 47.45 and 87.92 for SF-12v2. The SF-12v2 had larger standard deviations than the SF-36v2. No floor effects were detected; but ceiling effects were significant in the subscales of physical functioning （PF）, role-physical (RP), bodily pain (BP), social functioning (SF) and role emotion (RE). The SF-12v2 had higher ceiling effects (56.66%-68.32%) than the SF-36v2 (50.14%-63.87%). The exploratory factor analyses extracted two factors in both cases, representing physical (PCS) and mental health (MCS), respectively. The total variances explained by the common factors reached 64.05% for the SF-36v2 and 55.79% for the SF-12v2. The SF-12v2 PCS and MCS scores explained 91.0% and 80.3% of the total variances of those of the SF-36v2, respectively. The effect size of PCSs ranges from 0.78 to 2.77 in the subpopulations with different health conditions, compared with 0.00-0.57 for MSCs. The relative validity (RV) of PCS-12 to PCS-36 ranged from 0.89 to 0.94, compared with a MCS-12 to MCS-36 range of 0.60-0.75. Conclusion SF-12v2 is reliable and valid as a brief substitute version of SF-36v2 with acceptable responsiveness and equitable structure for assessing health related quality of life in the general population of Chengdu. But sub-scale scores were not recommended when using the SF-12v2 due to reduced precision.
