This study reports on an investigation of multi-date water-quality calibration algorithms under turbid inland water conditions using Landsat Thematic Mapper (TM) multispectral digital data. TM data and water-quality observations were obtained near-simultaneously in Green Bay, an embayment of Lake Michigan, and related using linear regression techniques. The high concentrations of terrestrially-derived suspended solids appear to be controlling the water volume reflectance. A simple exponential model using the band 3/band 1 (0-660μm/0-485μm) ratio was found to be a useful index for estimating total suspended solids and water transparency (i.e. Secchi disk depth) with an accuracy of ± 25 per cent over the calibration data range.