<jats:p> Classified-vehicle counts are a critical measure for forecasting the health of the roadway infrastructure and for planning future improvements to the transportation network. Balancing the cost of data collection with the fidelity of the measurements, length-based vehicle classification is one of the most common techniques used to collect classified-vehicle counts. Typically, the process of classifying vehicles by length uses a pair of detectors in a given lane to measure effective vehicle length. While the calculation is simple and seems well-defined, this paper demonstrates that small changes in the calculations can lead to large differences in performance during challenging conditions. In particular, most conventional calculations assume that acceleration can be ignored, and this assumption simply is not the case in congested traffic. As a result, many operating agencies are reluctant to deploy classification stations on roadways where traffic is frequently congested. This study examined six variations of the conventional calculation of vehicle length and developed a seventh that also estimated constant acceleration. The study then highlighted two of these approaches that worked well in extreme conditions on freeways for speeds as low as 15 mph. This range should be sufficient for most applications. Then, with empirically collected data, the authors found that the extreme events were uncommon and even the conventional method did quite well in stop-and-go traffic because the slower that traffic moved, the lower was the flow during that period. In any event, the key to success is the use of well-tuned detectors. </jats:p>