Multiparous species present a special set of problems and opportunities for study design and analysis. The present review reintroduces old concerns and raises new ones with empirical illustrations. Evidence is presented that litter effects are pervasive and persist into adulthood. Unaccounted for, they lead to spurious findings, inflate real effect sizes and produce false negatives. Furthermore, two-stage sampling, the practice of sampling only a subset of littermates from dams, can lower reliability, and therefore power, to unacceptable levels. In addition, the greater sensitivity offered by within-litter analyses over between-litter analyses is demonstrated. Statistical and experimental solutions are suggested and referenced. Surveys of recent developmental studies showed that the great majority do not attend to these issues, thereby casting doubt on the validity of their positive and negative findings. All developmentalists can strengthen their research by systematically addressing these concerns in study design and analysis.