This paper discusses various issues surrounding the use of imputation in the Household, Income and Labour Dynamics in Australia (HILDA) Survey and seeks a way towards an imputation strategy. While much of the discussion centres on the imputation for Wave 1, some consideration is also given to the strategy for later waves where it might impact on Wave 1. The various options for imputation are discussed and the recommended approach for the HILDA Survey draws on the experiences of other organisations with large-scale surveys, especially longitudinal surveys.