The possibility of using reliable behavioural parameters in toxicity determinations of molluscicides was assessed through the development of a method based on W.H.O. standard procedures of bioassying molluscicides and involving behavioural records by time-lapse cinematogrphy. B. glabrata adults (5 7/8 ± 1/8 whorls) were subjected to different sublethal doses of copper sulfate during 24 hours and then transferred to deionized distilled water for recovery; from those records it was possible to compute: (a) frequency of climbs to surface, (b) frequency of crawlings out of water and (c) proportion of snails on the upper, middle and botton thirds of the test containers. The Litchfield-Wilcoxon test was employed in determining a reference value (called "concentration of behavioural effect of 50%" of CBE50) in relation to each parameter. The indices thus obtained - (a) 0,010, (b) 0,006 and (c) 0,029 ppm of copper - showed the freasibility of systematic uses of behavioural criteria of toxicity and also proved capable of detecting toxic effects of the product under concentrations much lower than those obtained from conventional lethality determinations. The data also showed an effect/log dose lienar relationship for all the parameters considered and revealed changes in snail activity as a consequence of the daily light cycle. Although the clarification of the ethological aspects involved in the control of schistosome hsot snails depends on the analysis of the relationships between the snail and its natural habitat, laboratory studies, carried out with accurate measuremente of the parameters related to protective modes of behaviour, can be of great value as well.