For content questions, Wolof uses interrogative proforms without additional particles or case marking. Question words do not in themselves indicate their syntactic functions. Neither – due to the general fronting rule for question words – is the syntactic function of interrogative pronouns marked by their position, as is the case for arguments in declarative sentences in this SVO language. This role is taken on by verb inflection and derivational verb suffixes, which both contribute to specifying the argument structure. These pronouns are neatly organized into two parallel sets according to a general pattern whereby a common interrogative morpheme combines with different noun class markers. The two sets of interrogative pronouns are composed of the same class markers with either an -u or an -an formative, each one triggering the use of a different conjugational paradigm. Through its use of noun class markers, Wolof provides a neat system for content question words, paralleling the one for noun modifiers and including the two adverbial classifiers for space and manner. The ‘who?’ interrogatives are clearly formed with the (singular and plural) class markers for ‘person, human being’, the only members of these two paired classes. The interrogative pronoun for ‘what?’ uses one of the possible class markers for the nowadays atypical word THING, in accordance with the generalized use of this weakly productive class l- for all pronouns of the THING family. As is often the case, the ‘why?’ interrogative is a periphrasis, using a verb (‘what causes?’) and presenting a remarkable negative counterpart ‘what forbids?’.The latter also yields a grammaticalized form functioning as a question particle for ‘why not?’. Beside the use of other stems for ‘when?’ and ‘how many?’, what is at first puzzling in Wolof is the existence of the two parallel sets of interrogative pronouns made of the same class markers but with two different suffixes. Since they are not in complementary distribution, these two sets seem to be redundant. However, the proposed reconstruction for the atypical locative question word ‘ana’ sheds new light on the origin of this dual system. The two markers used to form these interrogative pronouns correspond to (1) a spatial deictic -u used for suspensive pronouns, indicating that the designated entity is not spatially localized: the question words formed with this marker appear to be indefinite relative pronouns that have grammaticalized into interrogative pronouns; and (2) an interrogative stem -an, whose origin as a former locative question word is revealed by the study of this irregular interrogative (locative) predicate. By contrast with -u, -an appears to be the only dedicated interrogative marker for content questions in Wolof.