In this paper, three essential questions are raised concerning the simulation study of the Japanese krill fishery carried out by Butterworth (1988). Firstly, the simulation study used krill distribution parameter values based on the FIBEX results, whereas Japanese vessels often operate on layers which are much larger than the swarms detected during FIBEX surveys. Since very little searching time is spent on layers, the utility of searching time data appears doubtful. Secondly, the utility of indices based on the sum of primary and secondary searching times (PST+SST) is also doubtful because SSTs are markedly dependent upon the type of products being produced, and the demand for each product varies substantially from one season to the next. Finally, attention is drawn to the unrealistic behaviour of the simulation model with respect to the values used for the minimum catch rate per overall elapsed time. It is suggested that an improvement in our knowledge of krill distribution is necessary before the routine implementation of searching time data collection is considered.