Previous studies of the latencies for the detection of inspiratory resistive (R) and elastic (E) loads revealed that R loads were detected earlier in the loaded breath than E loads (Respir. Physiol. 34: 267--77, 1978). These results suggested that the load information generated by R and E loads have different temporal patterns. We hypothesized that these differences might provide a mechanism for subjects to accurately discriminate between R and E loads and secondly, that the R and E load sensations perceived by subjects should have different temporal characteristics. To test these hypotheses we studied six healthy subjects in a two-part study. In this first, two levels of R and E loads (near threshold and 4--5 times threshold) were randomly presented for single inspirations separated by 3--6 unloaded breaths for a total of 10 presentations each. Subjects indicated detection and type of load perceived by pressing R or E marker buttons. In the second part, using the same loading protocol, subjects squeezed a hand grip dynamometer to express the pattern of perceived load sensation. Results indicate that subjects cannot discriminate between near-threshold R and E load but can discriminate between R and E loads 4--5 times threshold with considerable accuracy. Mean grip responses show that perceived load sensations arise earlier and reach maximum values sooner during R loaded breaths. Grip responses are consistent with the previously reported differences in R and E detection latencies and support the hypothesis that load discrimination is mediated by differences in the temporal pattern of load information generated during R and E loaded breaths.