Examples of effective growth-and-yield model implementation and use are documented throughout the forestry world. However, in the perspective of sustainable forestry, comparing the development of management requirements with growth-and-yield research backgrounds, some doubts arise concerning basic research hypotheses. The conflict between the necessary simplification implied in models development and the essential complex nature of forests plays a key role in the issue. Furthermore, present modelling concepts stem from a traditional framework relying at the fundamental perception level on the enforcement of external regulations on the forest, having some predetermined optimal structure as a goal. If tomorrow's strategic planning places in the forefront of the framework Oldeman's paradox “objectives are best achieved by first forgetting them in favour of forest survival and complexity maintenance,” we will really face quite a sharp concept change. Yet, it is important to clearly focus the role of models as potential adjuncts to the professional judgement and wisdom of foresters given their rational scientific character as learning, communication and screening tools. What is attainable is not the high precision of quantitative predictions, at least as far as complex biological systems such as forests are concerned. Compact representations (e.g., minimal models and “modeles simples”) with few state variables and a biologically-driven content may provide reasonable indications on future states over a wide range of conditions, contributing to operational forest management guidelines within an adaptive framework.