Nonidentical development of bilateral traits due to disturbing genetic or developmental factors is called fluctuating asymmetry (FA) if such deviations are continuously distributed. Fluctuating asymmetry is believed to be a reliable indicator of the fitness and welfare of an animal. Despite an increasing body of research, the link between FA and animal performance or welfare is reported to be inconsistent, possibly, among other reasons, due to inaccurate measuring protocols or incorrect statistical analyses. This paper reviews problems of interpreting FA results in poultry and provides guidelines for the measurement and analysis of FA, applied to broilers. A wide range of morphological traits were measured by 7 different techniques (ranging from measurements on living broilers or intact carcasses to X-rays, bones, and digital images) and evaluated for their applicability to estimate FA. Following 4 selection criteria (significant FA, absence of directional asymmetry or antisymmetry, absence of between-trait correlation in signed FA values, and high signal-to-noise ratio), from 3 to 14 measurements per method were found suitable for estimating the degree of FA. The accuracy of FA estimates was positively related to the complexity and time investment of the measuring method. In addition, our study clearly shows the importance of securing adequate statistical power when designing FA studies. Repeatability analyses of FA estimates indicated the need for larger sample sizes, more repeated measurements, or both, than are commonly used in FA studies.