Using revealed preference subscriber level data from a survey which represents the urban population of Turkey, and operator level data concerning the period of 2004-2009 obtained from the regulatory authorities and the operators; this paper employs the multinomial and conditional logit methods to examine the operator choice decisions of subscribers in the Turkish mobile market. The paper particularly focuses on two issues. Firstly, it addresses the relationship between the mobile operator choice and the subscriber bases and average prices of the operators (size effects and price mediated effects, respectively). Then, it identifies the effect of the other households' choices on the operator choice of the respondents (social network effects). The results are in line with the theory that the prices and the subscriber base-and to a greater extent, the household effects are important factors taken into account during the subscription decisions. The results also indicate that, compared with the period in which prices were loosely regulated, the estimated effects of prices and installed bases for the whole observed period are almost halved, while the effect of household network roughly keeps its level. These results make the focusing on narrower sub-networks i.e. calling clubs or social networks, an appealing strategy for the operators with relatively smaller networks. The paper suggests that the regulatory authorities should consider the possible sources of network effects in their market power analysis and future regulations. Meanwhile, the policy makers should be careful regarding the network quality regulations since the competition in Turkish mobile market seems to take the form of an intensified quality competition. (C) 2013 Elsevier Ltd. All rights reserved.