Abstract Typing has become a pervasive mode of language production worldwide, with keyboards fully integrated in a large part of many daily activities. The bulk of the literature on typing expertise concerns highly trained professional touch-typists, but contemporary typing skills mostly result from unconstrained sustained practice. We measured the typing performance of a large cohort of 1301 university students through an online platform and followed a preregistered plan to analyze performance distributions, practice factors, and cognitive variables. The results suggest that the standard model with a sharp distinction between novice and expert typists may be inaccurate to account for the performance of the current generation of young typists. More generally, this study shows how the mere frequent use of a new tool can lead to the incidental development of high expertise. Significance statement Typing has become a pervasive mode of language production worldwide, with keyboards fully integrated into many daily activities. Many people, including university students, spend several hours a day typing. Such intensive practice may lead to high levels of achievement, perhaps comparable with those that professional typists had before the advent of personal computers, despite the fact that contemporary typing often relies on informal learning and accommodates a greater range of typing habits. Our preregistered study aimed at characterizing the more variable expertise currently prevalent in university students, by combining two complementary approaches to the study of expertise. The first focuses on identifying the habits that are associated with proficient performance. The second focuses on identifying the underlying cognitive processes that might differ between the most and least proficient individuals. The results show that using a keyboard frequently can, by itself, lead to the development of high expertise, and that the difference between the least and the most proficient typists is more quantitative than qualitative. The available database provides a useful benchmark for future experimental research on typing. Introduction The acquisition of typing expertise has seen a radical change in the last two or three decades, going from the formal systematic training of a very limited population of professionals to a variable, often disorderly and unconstrained process carried out by a wide portion of the general population. From the invention of typewriters at the end of the nineteenth century to roughly the end of the 1980s, typewriting was almost exclusively performed by trained professionals. These individuals acquired highly homogeneous skills through intense formal training, which consisted in learning strictly systematic finger-to-key mappings, dispensing from the need to look at their hands while typing, among other requirements. This population of so-called touch-typists has received considerable attention in the scientific literature (e.g. Cooper, 1983; Gentner, 1983a), with much of this research being based on chronometric measures of performance, such as number of words per minute (wpm), response times (RT, the time elapsing between a stimulus and the first keystroke), or inter-keystroke intervals (IKI, time elapsing between two keystrokes). Since the advent of personal computers and their progressive dissemination from the 1990s onwards, an ever-increasing population has regular access to keyboards. Typing skills have become more widespread but also more variable (Feit et al., 2016) as there might be more than one way to speed up a typist (Behmer & Crump, 2016). For many typists, high levels of typing performance are achieved through unconstrained sustained practice or experience (Grabowski, 2008). This is most probably the case in France, for example, where typing is still only alluded to in school curricula (French Ministry of National Education, 2021). More generally, the importance given to typing in academic curricula varies substantially across countries; for instance, in the UK, the USA, or Norway, typing is a central aspect of learning to read and write (Genlott & Gronlund, 2013; Trageton, 2005). A thorough assessment is thus required to understand what characterizes the range of typing expertise occurring in the current population of twenty-first century typists. Two complementary perspectives can be taken on this issue (see Fig. 1). One focuses on characterizing the habits that lead to, or, minimally, are associated with, proficient typing skills. The other focuses on identifying underlying cognitive processes that might differ between the most and least proficient individuals. Our goal in the current study was to combine these two approaches, in order to determine how various practice habits and cognitive factors are related to the level of achievement in typing (Fig. 1).