Accurate estimates of aerosol refractive index (RI) are critical for modeling aerosol-radiation interaction, yet this information is limited for ambient organic aerosols, leading to large uncertainties in estimating aerosol radiative effects. We present a semi-empirical model that predicts the real RI n of organic aerosol material from its widely measured oxygen-to-carbon (O:C) and hydrogen-to-carbon (H:C) elemental ratios. The model was based on the theoretical framework of Lorenz-Lorentz equation and trained with n-values at 589

𝐴𝐴 nm (𝐴𝐴589nm ) of 160 pure compounds. The predictions can be expanded to predict n-values in a wide spectrum between 300 and 1,200 nm. The model was validated with newly measured and literature datasets of n-values for laboratory secondary organic aerosol (SOA) materials. Uncertainties 𝐴𝐴 of 𝐴𝐴589nm predictions for all SOA samples are within

𝐴𝐴 ±5%. The model suggests𝐴𝐴 that 𝐴𝐴589nm -values of organic aerosols may vary within a relatively small range for typical O:C and H:C values observed in the atmosphere. Plain Language Summary Atmospheric aerosol particles play an important role in affecting the climate by interacting with radiation and water. However, we have limited knowledge of the optical properties of atmospheric organic aerosols, which make up a large fraction of sub-micrometer aerosol particle mass. One of the challenges is that the RI, that is, the intrinsic optical constant of organic aerosol (OA) material, is poorly constrained. The lack of knowledge on the RI of organic aerosols can cause large uncertainties in estimating their optical properties and radiative effects on climate. To address this knowledge gap, a semi-empirical model is developed and validated that predicts the real RI of OA material based on the widely measured bulk chemical composition in laboratory and field studies. The model predictions suggest that the RI of typical ambient organic aerosols may have relatively small changes, which supports a simplified representation of using a constant n-value for ambient OA in atmospheric models. Potential applications of the developed model also include improving remote sensing and in situ optical sizing of aerosols.