All environments in nature fluctuate, and a globally common mode of fluctuation is cyclical: daily, tidal, seasonal, and even decadal. Organisms must adapt the scheduling of their life cycles-or 'life-histories'-to fit the cycles of the physical environments in which they reside, as the right timing is crucial for their evolutionary fitness. At the seasonal timescale, the familiar timing of biological activity is called phenology, and shifts in phenology are now deemed as the most conspicuous consequence of global climate change. Case studies of phenological change, however, seem idiosyncratic and incongruent, highlighting how little we still understand about life-cycle adaptations to fundamental temporal cycles in the environment. My doctoral work has focused on tackling this urgent issue using a three-pronged approach of theory, fieldwork, and experiment. At the center of it, I formulated a mathematical approach to modelling and analyzing the evolutionary process of life-history timing, and how ecology (e.g. population dynamics or changes in the underlying environmental cycle) influences such evolutionary processes. Notably, this approach is scale-free in that it can be applied to daily, tidal, seasonal, multiannual cycles, or any period length in between, and is easily adjustable for any species of interest given simple parameterization. I showed the conceptual implications of the theory as it connects to the literatures of life-history evolution and population ecology. Then, I demonstrated the model framework's power by parameterizing it for a particular species, a marine intertidal copepod Tigriopus californicus, for which I determined data can be accrued at amounts and speeds that are rarely high for life-history studies of wild populations. I showed that the theoretical framework had surprisingly strong predictive power in explaining patterns of life-history variation in wild populations. To more rigorously test the mean / variance dynamics of evolution of life-history traits, I conducted a long-term selection experiment with T. californicus populations in the laboratory and found that cycle periodicity is an exceedingly strong driver of intrapopulation trait variance, even more so than random noise in the environment. This result adds a novel contribution to the broadly relevant topic of how phenotypic variation is maintained in natural populations. To strengthen the bridge between theory and empirical work, I conducted an agent-based model that corroborated that slower (higher periodicity) environments increase life history variance in dynamic populations, and that the trajectory of trait distributions through transient and long-term phases is heavily influenced by environmental periodicity. Altogether, my work highlights that the typical stochastic characterizations of fluctuating environments--largely enforced by mathematical convenience--are insufficient in predicting and analyzing the evolution of life histories in dynamic populations which often reside in cyclically fluctuating environments.