In the present work it is demonstrated whether the acoustic distance calculated previously (Molina-García 2019) between [θ] and [s] based on their spellings -〈s〉 and 〈z, ce, ci〉- of Andalusian speakers is perceived by the speech community. Next, it is proved if it is possible to predict when [θ] realizations come from lexical series with 〈s〉 (CASA, 'house') and when from series with 〈z, ce, ci〉 (CAZA, 'hunt'). Finally, the social perception of this process is described. To this end, the performances of a 23-year-old informant with university studies and a 56-year-old informant with basic studies were recorded from the reading of 30 minimal pairs [s] - [θ]. 54 judges had to predict which word they had read. The results show that acoustic distance is perceived, that the informant's achievements do not differentiate with the vernacular pattern (ceceo), and that acoustic distance enjoys social prestige.