The alleged absence of musical pitch information for pure-tone frequencies above about 5 kHz is often cited as constituting evidence for temporal coding of pure-tone pitch below 5 kHz. However, the experimental data on which this allegation is based are both meager and equivocal. In this study, pure-tone pitch perception for frequencies above 10 kHz was investigated using three paradigms: open set melody recognition by naive observers, and melodic dictation and musical interval adjustment by musicians. The results indicate that significant musical-pitch information is available for frequencies above 10 kHz, although it is severely degraded relative to information at low frequencies. We have also investigated another pitch phenomenon which is assumed to be temporally based, the pitch differences associated with amplitude differences between the two components of unresolved two-component complex tones. This pitch information is also significant, but degraded, for frequencies above 10 kHz. The degree of degradation of both types of pitch information at high frequencies is consistent with the relative increase in pure-tone frequency DL's at high frequencies and is thus not incompatible with certain temporal models. [Supported by NINCDS.]