The purpose of this article is to summarize the current literature on the accuracy and reliability of interval systems using data from previously published experimental studies that used either human observations of behavior or computer simulations. Although multiple comparison studies provided mathematical adjustments or modifications to interval systems, recommendations were often noted as impractical or inefficient for use by early intervention/early childhood special education (EI/ECSE) practitioners. Based on the review of previous comparisons, none of the interval systems were accurate estimates of frequency; thus, event recording should be used to measure counts of behavior in early childhood settings. Momentary time sampling (MTS) may be appropriate and feasible for estimating duration, but only if random error is considered; partial (PIR) and whole interval recording (WIR) systems are inappropriate unless a statistical correction procedure is devised. Recommendations for using interval-based systems in applied research with young children with disabilities are provided.