A photodiode sensor has been designed and flown in an experiment to measure the broadband optical power of lightning. Several such sensors were launched on one rocket‐borne payload (apogee over 400 km) and four balloon‐borne payloads (float altitudes over 30 km) during the Wave Induced Particle Precipitation (WIPP) campaign (Kintner et al., 1987) at Wallops Island during the summer of 1987. Sensors aloft measured the optical power of events occurring below the payloads along with the waveforms of electric field transients. The same transients were recorded simultaneously by a ground‐based VLF receiver with a magnetic (loop) antenna. Rocket and balloon sensors detected a great majority of lightning flashes also recorded by the State University of New York at Albany (SUNYA) lightning locating network. Many other signatures which were similar to those of lightning but were not recorded by SUNYA were also detected. Overall during the 10‐min flight, more than 500 lightning‐related events were detected in the (wide) field of view of the rocket‐borne photodiode sensor. Among these is a class of about 23 events all having an anomalous signature, with obvious clustering of optical impulses or continuous emissions, and resulting durations of several hundred milliseconds. Such durations are much longer than typical for the lightning‐related events recorded at the rocket, which are more frequent overall. Every anomalous optical event (AOE) was accompanied by broadband VLF signals of a distinctive character, signals which were received on the balloon, the rocket, and the ground. No such AOEs were ever detected by the balloon optical photodiode sensor even when the data suggested that a signal from the AOEs also should have been detectable at the balloon. In fact, no such unusually long‐enduring optical events were ever detected by a balloon photodiode sensor on any of six separate flights in 1987 and 1988 under similar experimental conditions. Responses excluded so definitively from detection at the balloons are consistent with events originating primarily at altitudes above the 30 km balloon altitude. In considering possible sources above 30 km we find that the AOEs do not seem to resemble other natural optical phenomena, such as meteors which bum up well above 30 km in the mid‐latitude atmosphere.