Uranium isotope ratios have been determined for the fine-grained detrital fraction of Pleistocene Wilkawillina valley-fill sediments, four local Proterozoic bedrock samples and fine-grained aeolian material from a sand dune deposit of the Flinders Ranges, South Australia. The aim was to quantify the comminution age, i.e. the time elapsed since physical weathering of the bedrock, and residence time of the valley-fill sediments and to place tighter constraints on input parameters for the comminution age calculation. Despite using two independent approaches for determination of the recoil lost fraction of 234U from the sediment (weighted geometric and surface area estimates), samples fail to produce realistic comminution ages and hence, residence times. The issues involved in the ability to determine sediment comminution ages are discussed. The (234U/238U) activity ratio of the local bedrock is not in secular equilibrium, despite the bedrock being much older than 1Ma, i.e. the timeframe for 234U and 238U to reach secular equilibrium in a closed system. Using the average Flinders Ranges bedrock (234U/238U) ratio instead of an assumed (234U/238U) activity ratio of unity for the source would significantly reduce calculated residence times. This result warrants concern for future studies using the comminution approach for which a secular equilibrium source (234U/238U) activity ratio is assumed. Significant input of aeolian material may modify the measured (234U/238U) activity ratios. Such input may be more tightly constrained in future studies using rare earth element and radiogenic isotopic data. Future comminution studies would benefit from further consideration of the importance of 1) leaching lost 234U from source rock and bulk sediment samples, 2) wind deposition of fine-grained material and 3) the appropriateness and robustness of sample pre-treatment procedures.