This toxicity, or lack of it, has seemed to be bestowed upon fluoro-olefins in a random fashion. Attempts to correlate trends in toxicity with increasing fluorine substitution by ascribing olefin hydrolysis to hydrogen fluoride have not been entirely satisfactory. For example, perfluorobutene-2 and perfluoroisobutylene are isomeric, but although the toxicity of perfluoroisobutylene is well known, perfluorobutene-2 is not considered especially dangerous. Also, chlorotrifluoro-ethylene is about forty times more toxic than tetrafluoro-ethylene, and perfluoroisobutylene is about ten times as toxic as phosgene although both would be expected to be similar if hydrolysis were the only causative agent. One particularly interesting finding is the three-fold higher toxicity of trans-2,3-dichlorohexafluorobutene-2 over that for the cis isomer5.Little systematic attention has been directed towards determining comparative toxicities, for example LC50 or ALC concentrations, for this type of compound, but we have collected some of the available data in Table 1. It is clear from descriptions of physiological effects of these olefins that they do not act by a simple physical mechanism but rather by some biochemical interaction. Central nervous system involvement is often observed, particularly when these compounds are examined for possible anaesthesia potential. Further, of the few olefins shown to have high inhalation toxicity and that were also introduced in other ways, the toxicity was high for intravenous injection and also high for oral administration if in lipophilic solutions1. It seems that there is no reason to expect that olefin hydrolysis is the sole cause of toxicity, although its enhancement of the overall effect can be substantial.Olefin LC5o, p.p.m. ALC, p.p.m. (4 h exposure)CH2 = CHF 800,000 CH2 = CF2 128,000CF2 = CF2 40,000 CH2 = CC12 32,000CHC1=CC12 8,000 CC12 = CC12 4,000CF3CF=CF2 3,000 (3,240 to 13,365; 2h)CF2 = CFC1 1,000 (7,560; 2h)CF2 = CC12 1,000 CF3CF=CHCF3 200CF3CC1=CC1CF3 100 trans (61; 1 h)cis (179; 1 h) (CF3)2C=CF2 0.76Chiefly from refs. 4,9 and 10, LC50, lethal concentration for 50% of animals, 4 h exposure. ALC, approximate lethal concentration, 4 h exposure. A striking correlation apparently exists between fluoro-olefins and their toxicological properties which has not been previously realized: the toxicity of a halogenated olefin is directly proportional to the reactivity of that olefin to nucleo-philes.We shall not discuss in detail the postulates of nucleophilic substitution at unsaturated fluorocarbons or of fluorcarbanion theory. Nevertheless, fluorinated olefins are rather reactive towards nucleophilic reagents to yield /mostly substituted products and, more rarely, addition products. One may suggest a carbanion intermediate for these reactions although its descrite existence in some instances is not beyond question. This carbanion concept is useful as a predictor of both reactivity and products6. Generally, a fluorine moiety activates a double bond towards nucleophilic attack and the order of reactivity increases with intermediate carbanion stability, that is, 3 >2 >r. For example, we find the following order of nucleophilic susceptibility for perfluorinated olefins: (CF3)2C=CF2 > CF3CF=CF2 > CF2 = CF2This ranking is identical to that of the comparative toxicities of these olefins. Other halogen substituents affect olefin reactivity, usually enhancing it, although there is little information on the comparative kinetics of various fluorinated olefins towards nucleophiles. Most observations of relative nucleophilic susceptibility are deduced from the severity of experimental conditions needed to prepare vinyl ethers by reaction of olefins with alkoxides or alcohols.In at least one instance, the relative reactivity of a cis: trans isomer pair was determined. Park and Cook7 found that for the reaction of cis and rm^-2,3-dichlorohexafluorobutene-2 with methoxide ion at 25 C, the trans isomer was nearly four times as reactive as the cis isomer. Raventos and Lemon5 have found that the trans isomer of this olefin is about three times as toxic as the cis isomer. Surely these are more than coincidental relationships. Alkylating agents-compounds sensitive to nucleophiles- are a well known class of compounds possessing high biological activity8. That fluorinated olefins may also be alkylating agents in biological systems must also be considered. Further, it is possible that this is the principal reason for the high toxi-city found in many members of this class. Of course, halide hydrolysis is a concurrent toxic mechanism and is doubtlessly responsible for lung oedema and associated disorders. We feel, however, that this hydrolysis is obscuring what is likely to be the principal toxic mode of activity of these compounds. The warning is clear: whenever one works with a poly-fluorinated olefin of high nucleophilic susceptibility, and in the absence of specific toxicological information, extraordinary safety precautions are indicated.