The last exposure to sunlight of detrital feldspars and quartz grains in Quaternary sediments can be dated directly by thermoluminescence (TL) methods, but the effectiveness of zeroing of the light-sensitive TL signal by daylight varies widely, depending upon details of the sedimentation processes. To document this variability and to help develop objective criteria useful for future sample selection and laboratory procedures, I have collected and analyzed several eolian, fluvial, marine deltaic, and lacustrine samples from less-than-100-year-old deposits as well as suspensions from sediment plumes. Also, underwater spectral irradiance was measured during recovery of suspensions. TL properties of polymineralic (feldspar dominated) 2--4 μm and 4--11 μm size fractions were measured in the laboratory. Unexpectedly, some light-sensitive TL was found to remain in all samples. However, the light-sensitive TL was more effectively zeroed in some samples (a proximal loess, a sandy braided-channel deposit, tidal flat and upper-delta-front-muds, and thin clayey lacustrine laminae) than in others (a silty braided-channel deposit, submarine fan muds, and silt-bearing lacustrine laminae). For the first group the minimum relict paleodose obtained with the partial-bleach technique is about 0.4¿0.3 Gy, corresponding to a relict TL age estimate of about 200 years for these sediments. Thus similar sediments (excluding fluvial deposits) as young as about 2 ka can be accurately and routinely dated without significant correction for a zero-point error if one uses the partial-bleach technique coupled with integrated optical-bleaching energy fluences of less than ≂1 J cm-2 and wavelengths only above 550 nm. Also, some simple optical-bleaching tests are reported that have a potential use in sample selection and in provenance studies of some surficial detrital sediments. Finally, a simple mathematical model is presented that helps account for the gross features of the observed bleaching-response curves. ¿ American Geophysical Union 1990