Theoretical incoherent scatter (IS) spectra from collisionless F region plasma are calculated taking into account the effect of negatively charged ions. Such ions can significantly change the composition of the local ionosphere when they are created artificially during active chemical release experiments. If IS radar observations are conducted upon such a modified plasma, the interpretation of the scattered radar signal will be complicated by the negative ions, which are not found in the natural (e.g., O+/e-) ionosphere. Specifically, the presence of heavy negative ions will broaden the IS spectrum, making the shoulder peaks sharper, and creating an additional small central hump. A radar spectral analysis program that assumes only the naturally occurring plasma constituents will erroneously attribute such spectral changes to changes in other plasma parameters, such as the temperature ratio Te/Ti. In order to detect heavy negative ions we fix the temperature structure of the ionosphere to a preevent average measurement and interpret any changes in spectral shape during the experiment as being caused by changes in composition, and not by changes in Te/Ti. The Millstone Hill 440-MHz IS radar was used to observe the spatial and temporal development of heavy negative ion plasma clouds created during four active chemical release experiments: the Ionospheric Modification Study (IMS) in 1983, the Space-Plasma Negative Ion Experiments (SPINEX 1 and 2) in 1984 and 1986 (all SF6 releases), and NICARE 1 in 1989 (CF3B release). Each experiment led to the creation of a region of heavy, negatively charged ions. Concentrations of 10-40% SF6- (146 amu) are reliably detected in the SPINEX 1, SPINEX 2 and IMS data sets. An average uncertainty of ¿10% SF6- is present in all three experiments. Concentrations of 30% Br- (80 amu) are further detected in the NICARE 1 release, with uncertainties of ¿4%. ¿ American Geophysical Union 1992