Following on from the apparent failure to satisfactorily assess the status of the C. gunnari population in Subarea 48.3 at WG-FSA-91 using VPA, attempts were made to re-work the analysis using Laurec-Shepherd and ADAPT tuning techniques, from 1991 back to 1977. The predicted age structure, dominated in recent years by the 1987 year class (1 year olds in 1988), was quite robust, despite the use of various combinations of survey and cpue indices for tuning. According to the VPA the population in 1991/92 was composed of a large proportion of 5 year olds, which was not observed during the survey on Falklands Protector in January 1992. Breakdown in the credibility of the VPA results in most recent years is attributed to the invalid assumption of constant M and contradictions in the input data. A conservative approach to management for 1992/93 is recommended, based on the results of surveys by Falklands Protector in 1990/91 and 1991/92.