Ocean surface algal blooms (OSABs) are an emerging and progressively more important phenomenon in many regional waters. Current techniques perhaps delineate the spatial distributions of OSABs from the above-atmosphere spectral radiances or Raleigh corrected radiances, so far optical spectral characterization of the OSABs has been impossible due to the inability of the atmospheric correction algorithm to account for their enhanced radiances at the near-infrared (NIR) bands. An innovative algorithm, referred as the “OSABT”, is developed to provide essential data for detection and optical spectral characterization of OSABs in coastal and open ocean waters. The algorithm is applied to MODIS-Aqua imagery from the Arabian Sea and its results are systematically studied with regard to the spectral characteristics of OSABs, its evolution stages, sub-pixel variability, and atmospheric correction issues. The effectiveness of the OSABT is tested for several other regional waters on MODIS-Aqua data, where spatial bloom characteristics are previously known, and their optical spectra are featured with enhanced radiance at the NIR bands and reduced radiance at the red bands. The spectral signatures differ for each OSABs, and correlate to their evolution stages as well. These results support future applicability of this algorithm to derive the water-leaving radiance and new indices from the OSAB pixels for qualitative and quantitative assessment of the optical and biological OSAB properties as well as their associated ecological activities (e.g., growth, aggregation, density and impacts). The advantage of the OSABT is that it can be applied rather generally to other satellite-based sensors with no short-wave infrared bands, provided that suitable radiometric calibration coefficients are determined for using the OSABT.