There are many factors that contribute to successful second language learning, including exposure to the language, explicit instruction on grammar and multilingualism. In this online study, participants’ ability to acquire a complex artificial language via cross-situational statistical learning (CSSL) was assessed. Participants from diverse linguistic backgrounds were recruited using social media. Participants’ age of acquisition, exposure, and proficiency in their languages were measured in a questionnaire. Then, participants were divided into two groups; in the explicit instruction group, participants received direct information on what parts of the stimuli they should pay attention to and why, while in the implicit instruction condition, they were simply told to learn the language. The language to be learnt allowed subject-object-verb as well as object-subject-verb order, and the subject and the object were case marked by postpositions. Participants were exposed to five-word long transitive sentences and had to decide which of two animations depicts the meaning of the utterances for around twenty minutes. Vocabulary tests were incorporated into the learning trials, and participants were tested on their knowledge of the syntactic word order of the language afterwards. Preliminary results showed that participants learnt the language regardless of the instruction and their linguistic background. The findings of this study contribute to our understanding of what factors influence statistical learning of a complex artificial language and suggest that CSSL not only endow learners with the ability to recognise correct syntactic structures but also to produce them.