The adaptive landscape has been suggested as a potential conceptual bridge between phenotypic evolution on generational to macroevolutionary timescales. However, this potential remains largely untapped due to a limited understanding of how the adaptive landscape changes across time. A fixed adaptive landscape is usually assumed in microevolutionary studies, and stabilizing selection around a static peak is a commonly evoked explanation for stasis in the fossil record, while macroevolutionary change is often claimed to be associated with sudden movements of peaks on the adaptive landscape. We assessed the dynamics of the adaptive landscape across various timescales by analyzing different evolutionary time series using diverse multivariate models of evolution. First, we examined whether a human-induced decrease in river waterflow affected the optimal body mass of a salmon population over a few decades. Second, we explored whether changes in oxygen and carbon isotopes (proxies for temperature and possibly nutrient availability) affected the optimal size of a species of coccolithophore across a hundred thousand years in the late Albian. Finally, we analyzed the extent to which oxygen and carbon isotope variations affected the optimum size in a planktic foraminifera lineage over a few million years during the Miocene. Results support a dynamical adaptive landscape in two of the datasets covering micro- and macroevolutionary timescales, meaning that the salmon population as well as the foraminifera lineage had to constantly readapt to environmental changes in the positions of adaptive peaks. The non-conclusive results observed for the coccolith dataset show the complex topography of the adaptive landscape. Although the rate of adaptation and evolution varies among the three lineages, adaptive landscapes may be more dynamic than often assumed and advocated. Multivariate analyses of time series may provide valuable insight into how changes in the adaptive landscape lead to evolutionary changes in phenotypes across micro to macroevolutionary timescales.