This study examined the effects of some procedure variables on the emergence of contextual control over stimuli classes. Two experiments were conducted each with four college students. In both experiments, during training, contextual stimuli {Xl and X2) were shown with only a few of the stimuli later to be used as conditional stimulus on tests of contextual control over conditional relations. In Experiment 1, Phase 1 consisted of contextual control training over conditional relation (train XAB) and simple conditional relations training (train BC and DA relations), followed (Phase 2) by tests of contextual control over conditional relations (tests XCD, XCA. XDC, XBD, XAC, XDB and XBA) and symmetrical relations (CB and AD). In Phase 3, the XY and XZ training were employed, followed (Phase 4) by equivalence (ZY and YZ) and xymmetrical (ZX e YX) tests. In Experiment 2, the same procedures used in Experiment 1 were employed, but in reversed order: the training and tests of contextual control over conditional relations took place after training (XY and XZ) and tests (ZY, YZ, ZX, YX), and DA and BC training. Both experiments ended (Phase 5) with the testing of transfer of contextual control between equivalence stimulos over the conditional relations previously trained and tested. Three subjects, one for Experiment 1 and two for Experiment 2, showed emergence of contextual control over equivalence relations without prior exposure to any other tests. Three other subjects (one for Experiment 1 and two for Experiment 2) achieved the same results only after the emergence of contextual control over conditional relations defined by lower nodal number. These results suggested that acquisition of experimental history with conditional responding by exposure to simple conditional discriminations (train and test) before training and tests of contextual control over conditional relations, can be an important procedure variable to the emergence of this kind of stimulus control over equivalence stimuli classes.Este estudo investigou efeitos de variáveis de procedimento na emergĕncia do controle contextual sobre classes de estímulos. No Experimento 1, Fase 1. ocorreram os treinos do controle contextual (treino XAB) e das relaçōes condicionais BC e DA, seguidos (Fase 2) pelos testes do controle contextual sobre relaçōes condicionais. Na Fase 3, ocorreram os treinos XY e XZ, com teste posterior (Fase 4) das relaçōes ZY, YZ, ZX e YX. No Experimento 2 ocorreu inversâo das fases: após os treinos XY e XZ, os testes ZY, YZ, ZX e YX, e os treinos DA e BC, ocorreram os treinos (XAB) e os testes do controle contextual. Na Fase 5, nos dois experimentos, avaliou-se se os estímulos Z e Y exerceriam o controle contextual exercido pelos estímulos X. Trĕs sujeitos, um no Experimento 1 e dois no Experimento 2, registraram a emĕrgencia do controle contextual sobre as relaçōes equivalentes sem a necessidade da exposiçâo previa aos demaie testes. Outros trĕs sujeitos (um do Experimento1 e dois do Experimento 2) registram tal resultado após a emĕrgencia do controle contextua! sobre as relaçōes equivalentes com menor número nodal. Sugere-se que o establecimento de uma história experimental com relaçōes condicionáis simples, precedendo aos treinos e aos testes do controle contextual, pode constituir-se em variável relevante na emĕrgencia desta modalidade de controle de estímulos.