Improving understanding and prediction of the potato (Solanum tuberosum) tuber size over the growing season is important due to its effects on crop price and marketing. Several models have been proposed to describe potato growth and development, but are based on short-term data and have little use for predicting yields or in-season management decisions. This analysis uses long-term data collected from 1979 to 1993 in central Wisconsin to describe growth and development of the Russet Burbank potato variety. This paper describes average number of potato tubers per plant and tuber length as influenced by thermal time and stem number per plant over 14 years. For each plant variable, data analysis uses multivariate techniques to fit a hierarchical logistic model with parameters potentially depending on stem number per plant. Analysis finds that the average number of potato tubers and average tuber length were affected by thermal time and stem number per plant. Estimated models are biologically relevant, provide an understanding of seasonal thermal variability and stem number per plant effects on average tuber set and growth, and can be used to describe yearly variation in average potato growth and development. Increased understanding of potato growth in response to thermal time and stem number per plant can improve management recommendations and predictions of crop economic value. Full article