The possibility to estimate ages and masses of Young Stellar Objects (YSOs) from their location in the Hertzsprung-Russell diagram (HRD) or a colormagnitude diagram provides a very important tool for the investigation of fundamental questions related to the processes of star formation and early stellar evolution.Age estimates are essential for studies of the temporal evolution of circumstellar material around YSOs and the conditions for planet formation.The characterization of the age distribution of the YSOs in a star forming region allows researchers to reconstruct the star formation history and provides important information on the fundamental question of whether star formation is a slow or a fast process.However,the reliability of these age measurements and the ability to detect possible age spreads in the stellar population of star forming regions are fundamentally limited by several factors.The variability of YSOs,unresolved binary components,and uncertainties in the calibrations of the stellar parameters cause uncertainties in the derived luminosities that are usually much larger than the typical photometry errors.Furthermore,the pre-main sequence evolution track of a YSO depends to some degree on the initial conditions and the details of its individual accretion history.I discuss how these observational and model uncertainties affect the derived isochronal ages,and demonstrate how neglecting or underestimating these uncertainties can easily lead to severe misinterpretations,gross overestimates of the age spread,and ill-based conclusions about the star formation history.These effects are illustrated by means of Monte-Carlo simulations of observed star clusters with realistic observational uncertainties.The most important points are as follows.First,the observed scatter in the HRD must not be confused with a genuine age spread,but is always just an upper limit to the true age spread.Second,histograms of isochronal ages naturally show a decreasing number of stars for ages above the median,a pattern that can be misinterpreted as an accelerating star formation rate.Third,it is emphasized that many star forming regions consist of several sub-groups,which often have different ages.If these distinct stellar populations cannot be disentangled (e.g.,due to projection effects) and the HRD of all stars in the region is used for an age analysis,it is very difficult (often impossible) to discern between the scenario of an extended period of star formation (i.e.a large age spread) and the alternative concept of a temporal sequence of several discrete star formation episodes.Considering these factors,most observations of star forming regions suggest that age spreads are usually smaller than the corresponding crossing times,supporting the scenario of fast and dynamic star formation.