Scale is highly detrimental to the surface quality of tinplate products and this problem is created during the hot rolling process. In this paper, a multinomial logit model is used to both quantify the influence of hot mill process parameters on tertiary scale formation and to identify how to optimize the running conditions at the hot mill, that is, to identify what base chemistries to use and what temperatures to run the hot mill at so that the percentage of coils produced with significant scale formation is kept as low as is required for meeting customer requirements. Principal component analysis is used to reduce the dimensionality of the dataset while retaining the majority of the variability in the process variables. It was found that a multinomial logit model containing these components was consistent with the process data and it was further determined from this model that the most significant variables were the temperature at which the steel strip was entering the finishing mill, together with the percentage of phosphorus and aluminum present within the slab entering the mill. More importantly, for process control it was found that to keep the rate of coils containing medium or high scale counts below 10%, the average rougher mill temperature should be kept at 1020 °C or less, or alternatively the aluminum content should be kept below 0.024% wt when running the hot mill at the mean values for all the other process variables. The multinomial logit model was capable of identifying many other optimal running conditions for the hot mill.