The supply of steel blocks with customised dimensions for clients within the metalworking industry requires the effective management of a large and highly variable stock of steel blocks. This paper proposes a cutting decision process for daily tailored orders of a real-life steel retailer. The problem is approached as a three-dimensional (3D) cutting and packing (C&P) problem with variable sized blocks and daily changing characteristics, which results in high computational complexity. A novel definition is proposed for this type of problem. Large time periods are considered, as one day cutting configuration has impact on the following days. The proposed approach considers the usable leftovers of the preceding days to the input stock of the next day. A first-fit decreasing (FFD) and best-fit decreasing (BFD) approaches are proposed to deal with the problem, in order to minimise the stock growth. The results show that the proposed algorithms are able to keep the stock size low and to generate a small number of stock pieces with low weight. [Received 1 May 2015; Revised 11 December 2015; Accepted 25 January 2016]