Objective The objective of the present study was to validate sequela prediction by the Maximal Injury Impairment Score (M-IIS), in comparison with the Functional Independence Measure (FIM) assessed at 1 year's follow-up of severe road crash victims. Methods The study population came from the ESPARR cohort: 178 victims (with Maximal Abbreviated Injury Scale ≥3) of road crashes in the Rhône administrative département of France, aged ≥16 years and with medical examination including FIM scoring 1 year post-accident. Two thresholds were tested for both scores. Firstly, the relation between FIM and M-IIS was assessed on logistic regression models adjusted on age and presence of complications at 1 year post-accident. The predictive capacity of M-IIS was expressed as its negative and positive predictive values, and considered as good when ≥80%. Results Sixty-three of the 178 adult subjects (mean age = 37.7 years; range = 16.1–82.9 yrs) showed post-accident complications. One-year sequela prediction on M-IIS was greater in head, spine and limb lesions, but limited to slight impairments (M-IIS = 1). There was a significant correlation between FIM and M-IIS, although “age” and “medical complications” were confounding factors on certain multivariate models. The predictive capacity of M-IIS was low, for whatever type of sequela. Conclusions M-IIS, in this severely injured population, failed to predict sequelae at 1 year as measured by the FIM, despite a good correlation between the two. Complications are to be taken into account in assessing the M-IIS's capacity to predict sequelae. Further evaluation will be needed on larger series or assessment of other indicators and measures of sequelae at 1 year, to obtain a robust tool to predict road crash sequelae.