An alternative to the current approach to freeway bottleneck capacity analysis is outlined: average time gaps (average time separations between the rear of a vehicle and the front of one following it), average passage times (average times for vehicles to pass a point), and lane flow distributions serve as intervening variables to link geometric, vehicle population, and driver population characteristics to capacity flows. The first phase of research to evaluate this approach is described. The goal was to investigate interrelationships between the intervening variables and to relate them to average flow per lane under capacity conditions. Two types of "capacity condition"—pre-queue flow and queue discharge flow—were considered. Traffic data for 18 extended data collection periods at 15 bottlenecks in the Minneapolis-St. Paul, Minnesota; San Diego, California; and Seattle, Washington, areas were analyzed to calculate average values of critical lane time gaps (where the critical lane is that with the highest flow rate), critical lane passage times, and critical lane flow ratios (ratio of critical lane flow to flow per lane) and to relate them to average flows for periods of pre-queue and queue discharge flow. Critical lane average time gaps were not significantly correlated with critical lane flow ratios but were significantly correlated with average passage times and with flow per lane in both pre-queue and queue discharge flow; however, critical lane flow ratios were significantly correlated with flow per lane in queue discharge flow only. There is a strong, near-linear negative relationship between critical lane flows and critical lane average time gaps in both pre-queue and queue discharge flow.