The pervasiveness of the English language in society and education in the Netherlands, as well as its status as online lingua franca, has caused concerns. English is a manifest aspect of oral youth language, as reflected in online written messages, but has in no way replaced the Dutch language. This paper presents a large-scale corpus analysis of computer-mediated communication by Dutch youths. These social media messages (392,169 words in total) were studied for the presence of code-mixing with English, in terms of amount and manner. They contained 7,528 English elements: (parts of) words, interjections, textisms (typical of ‘digi-talk’), phrases, and sentences. We argue that the concept of ‘manifold code-mixing’, consisting of four pathways – discourse framing, insertion, alternation, and integration – is necessary to truly comprehend the complexity and social meaning of code-mixing. These pathways relate to the SUPER-functions of textisms (speechlike, understandable, playful, expressive, reduced) and reveal Dutch youths’ high proficiency in English.