We introduce the concept of joint source coding, channel coding, and secrecy. In particular, we propose two practical joint schemes: the first one is based on error-correcting randomized arithmetic codes, while the second one employs turbo codes with compression, error protection, and securization capabilities. We provide simulation results on ideal binary data showing that the proposed schemes achieve satisfactory performance; they also eliminate the need for external compression and ciphering blocks with a significant potential computational advantage.