The present study was an attempt to investigate the acquisition of negation properties by Persian monolingual and Kurdish-Persian bilingual learners of English across different levels of language proficiency and within a generative framework. Generative models are generally concerned with issues such as universal grammar (UG), language transfer, and morphological variability in nonprimary language development. Hence, an attempt was made to test the claims and predictions made by a number of generative theories specifically FTFA, RDH, DA, MSIH, SSH, and MSBH. To do so, 180 Persian monolingual and Kurdish-Persian bilingual learners of English participated in the study. Based on the Oxford Placement Test, they were assigned to 3 levels of language proficiency. They, then, received a grammaticality judgment test and a translation task. The results revealed that there was, first, no significant difference between the performance of the monolingual and bilingual learners at each level of language proficiency, whereas the difference was statistically significant across the levels of proficiency, and second, no single generative theory can offer a comprehensive explanation about the whole process of L2-L3 negation acquisition. Indeed, language development occurs in an incremental manner as predicted by MSBH, and the resetting of TL properties is possible with increasing level of proficiency. The findings also confirmed that knowledge of an L2 does not play a significant role in the development of the interlanguage grammars of Kurdish-Persian bilinguals, and increased L2/L3 exposure and use lead to less language transfer and consequently approximation to native-like performance. Each of these findings has been discussed in the framework of the related theories tested in the study.