The objective of this paper is to suggest a methodology for the practical inversion of acoustic backscatter to obtain suspended mass concentration of marine flocs. Optical data collected using a LISST-100 at the same time as acoustic backscatter (ABS) measurements were made during a series of controlled flocculation laboratory experiments using kaolin ( MacDonald et al., 2013 ) showed a consistent peak in the particle size distribution at around 10–20 µm in 50 of the 75 experimental runs. This peak corresponds to the size of the strongest and most tightly bound of component floc structures referred to as ‘flocculi’ ( Van Leussen, 1994; Lee et al., 2012 ). Log-normal distributions fitted to these peaks accounted for between 30.1% and 90.7% of the total volume of particles (median 50.4%). The equivalent mean floc sizes identified from the FLOCView camera system were between 38 and 92 µm. It was hypothesised that the acoustic backscatter from the flocculated kaolin was occurring from the flocculi (and not the flocs per se ) and that the flocculi could be treated as elastic spheres with density and acoustic wave propagation speeds equal to bulk kaolin. The ABS signals were inverted, using the methodology suggested by Lee and Hanes (1995) , using the form function and total scattering cross section defined by Moate and Thorne (2009) for particles with a broad size distribution, to determine the particle size. Excluding the unflocculated runs the particle diameters were 15.5±4.9 µm; the particle size tended to increase with concentration and with the ‘level’ of flocculation. The viscous attenuation, which cannot be neglected for particles of this size, was also estimated from the ABS data and was found to be consistent with Urick's (1948) model. It is concluded that the flocculi model of acoustic backscatter is supported by laboratory measurements on a simple suspension of kaolin, and may offer a practical way forward for the interpretation of marine field data. [ABSTRACT FROM AUTHOR]