Many potential applications for terahertz technology have been suggested at on time or another. These fall broadly into two categories, those which require the use of terahertz frequencies by their nature and those which are currently being developed at lower frequencies - below 100 GHz, say - but which might benefit from the use of shorter wavelengths were suitable technology available. Prominent among applications in the former category are astronomy and remote sensing. Two examples of satellite radiometer systems currently in the planning stages are MASTER and SOPRANO. The improvements in device technology which would greatly benefit these systems include the provision of a good solid-state local oscillator source capable of delivering at least 1 mW of power. Failing this improvements in the design and efficiency of frequency multipliers would be highly desirable. The development of the SBV may be significant here. As far as the mixer is concerned, apart from reductions in conversion loss and noise temperature improved device mount structures which are more rugged and reproducible than the conventional `cats whisker' mount are urgently needed. If such mounts can be used for device integration so much the better.There are specific applications for one-off scientific instruments for a variety of measurement applications. One example is given of a system designed to measure the scattering from a single-particle scatter. These measurements being in the nature of low frequency scale modelling of infrared scattering.In the millimetre wave region, up to 100 GHz there are several highly commercial applications, in particular high frequency communications and automobile radar. The question is would these be better implemented at higher frequencies were suitable device technology available. The consensus of opinion among works in this field is that there is little advantage in such a move, however, recent advances in the new techniques of micromachining indicate that if, in particular, the source problem could be solved, there may indeed be practical advantages to be gained by the use of shorter wavelengths.