Despite intensive study of morphological effects using various on-line techniques such as masked priming, psycholinguistic accounts of morphological processing have not yet managed to grasp the complexity of the various kinds of relationships between words. We focus on three issues related to aspects of processing that have not been given much importance but can considerably influence the effects we observe. The first issue to be dealt with is the role of frequency of the lexical items used and particularly the role of prime-target relative frequency. Second, 'morphological' effects with nonwords (e.g. sportation-sport), which are very often interpreted as if the status of possible word did not exist. Finally, the role of a novel variable, pseudo-family size, reflecting the influence of formally related but morphologically unrelated word forms, providing evidence for interference/competition during the early stages of morphological processing. These factors suggest that the complex set of activation/inhibition related to the lexical environment of the word-to-be-identified should be taken into account, and that morphological processing models should try to introduce factors related to the paradigmatic structure of language.