During the winter of 1996, a heavy storm crossed the coasts of the Azores. A comparison of wave-exposed and wave-sheltered samples of the intertidal planktonic developing periwinkle Littorina striata (Mollusca: Caenogastropoda), collected before and after this storm revealed: (1) an overall difference in relative and absolute aperture area between wave-exposed and wave-sheltered samples; and (2) depending on the technique used a storm (MANOVA/CDA/1994 vs 1996 nested ANOVA) or no storm (1992, 1993, 1994, 1996, nested ANOVA) effect was noted. Preference should be given to the latter analysis as it is more correct in the present context. Nesting of the factor ‘year’ in the factor ‘storm’, allows detection of the true storm effect, disentangling it from other year exposure interactions, while the inclusion of more than one pre-storm sample allows differentiation of storm effects from ‘normal’ temporal variation.