Abstract. During the last 25 years the Danjon astrolabe has been used extensively for measuring the apparent solar diameter. Here we discuss some aspects of the results obtained with the CCD astrolabes of Antalya (Turkey) and Rio de Janeiro (Brazil), and with the visual astrolabes of CERGA (France) and Santiago (Chile). The most extended databases obtained with these astrolabes are those of CERGA and probably Santiago. However, concerning observed variations in time of the solar radius, the results of both stations are discrepant. While the CERGA results show a marginal radius variation in opposing phase with solar magnetic activity, those of Santiago give a significant variation in time, but in phase with solar activity. According to Chollet & Sinceac (1999) the apparent solar radius variation is identical for all astrolabe measures (except Santiago) and anticorrelated with the magnetic cycle of the Sun. However, we show here that according to published results, the astrolabes of Antalya and Santiago give similar radius variation. With respect to the results of Rio de Janeiro, they show strong internal inconsitencies due probably to instrumental modifications made during its observational campaign. After removing these inconsistencies, the results of Rio de Janeiro become almost identical to those of Santiago. The results of these astrolabes, Antalya, Rio de Janeiro and Santiago, are in agreement with recent results obtained with different observational techniques, which show also that the apparent solar radius would vary in phase with solar magnetic activity.