In pointing, throughput TP is used as a performance metric for the input device and operator. Based on the calculation of effective parameters (width We and amplitude Ae), TP should be independent of the speed-accuracy tradeoff. To examine the validity of TP and effective parameters for crossing actions, we conducted two experiments using two established crossing tasks. Our results demonstrate that applying effective parameters to Fitts’ law model improves the fit to the data for mixed biases in both tasks. Besides, we observed that effective parameters smoothed TPs across biases. However, unlike pointing, TP was observed to be unstable across IDs in one task, while was stable across IDs in the other task. Analyzing speed profiles showed that this was likely due to the fact that one of the tasks could be completed with a ballistic movement at low IDs, whereas this was impossible for the other task.