Although quantitative characterization of growth phenotypes is of key importance for the understanding of essential networks driving plant growth, the majority of growth-related genes are still being identified based on qualitative visual observations and/or single-endpoint quantitative measurements. We developed an in vitro growth imaging system (IGIS) to perform time-resolved analysis of rosette growth. In this system, Arabidopsis plants are grown in Petri dishes mounted on a rotating disk, and images of each plate are taken on an hourly basis. Automated image analysis was developed in order to obtain several growth-related parameters, such as projected rosette area, rosette relative growth rate, compactness and stockiness, over time. To illustrate the use of the platform and the resulting data, we present the results for the growth response of Col–0 plants subjected to three mild stress conditions. Although the reduction in rosette area was relatively similar at 19 days after stratification, the time-lapse analysis demonstrated that plants react differently to salt, osmotic and oxidative stress. The rosette area was altered at various time points during development, and leaf movement and shape parameters were also affected differently. We also used the IGIS to analyze in detail the growth behavior of mutants with enhanced leaf size. Analysis of several growth-related parameters over time in these mutants revealed several specificities in growth behavior, underlining the high complexity of leaf growth coordination. These results demonstrate that time-resolved imaging of in vitro rosette growth generates a better understanding of growth phenotypes than endpoint measurements.