In the Arabic-speaking world, sociopolitical upheaval, extended conflict and population displacement have triggered extensive contact between mutually intelligible varieties of the language. Notwithstanding these developments, Arabic sociolinguistic research on dialect contact settings remains limited to certain well-documented areas (e.g., Al-Wer 2020), with markedly less research targeting other locales believed to be highly propitious to convergent change, such as the long-term contact situation in Lebanon involving Lebanese and Palestinian Arabic (Fityan 1981; Hennessey 2011). Furthermore, few studies are embedded in a (comparative) variationist sociolinguistic framework (Owens 2013), and even fewer studies are articulated from a socio-historical perspective incorporating diachronic data sources with which to better understand the process of language change in Arabic (Owens 2013). Much previous research on Arabic dialects is also based on investigations of phonological variation (Al-Wer and de Jong 2018), with correspondingly less attention paid to (morpho-)syntactic variation (Choueiri 2019). The present study aims to address existing lacunae in the research literature by investigating the outcomes of dialect contact in Beirut between Palestinian Arabic (PA), the minority variety, and Lebanese Arabic (LA), the majority variety. Drawing on the framework of comparative variationist sociolinguistics (Poplack and Tagliamonte 2001) as well as research on dialect contact (Britain and Trudgill 2005), this study combines synchronic and diachronic data sources to compare three variables in LA and PA: a phonological variable, involving the word-medial raising of /a:/ to [e:] (e.g., [ka:n] alternating with [ke:n] 'he/it was'); and two morpho-syntactic variables: verbal negation and future temporal reference. The overarching aim of the research is to examine the extent to which PA shows evidence, as gauged from linguistic constraints on variant selection and variant repertoires, of becoming more structurally similar to LA in different linguistic components (Cheshire, Kerswill, and Williams 2005). The synchronic data come from 45 hours of spontaneous speech recorded in Beirut from 39 Palestinian and 27 Lebanese speakers stratified by age, sex, and level of education, generating 7,671 tokens representing the three targeted variables. A further 15,381 tokens of these three variables come from two diachronic datasets. The first is a sub-set of speech recordings from the Palestinian Oral History Archive, an online compendium of interviews with first-generation (older) Palestinians in Lebanon, recorded between the 1990s and early 2000s. The second diachronic dataset is the Lebanese Popular Theatre Corpus (LPTC), based on 34 televised plays dating from the 1960s and performed in colloquial LA. Results reveal that the [e:] variant, a stereotypical feature of LA, but not emblematic of PA spoken in Beirut (Hennessey 2011), is virtually absent from the speech of the older Palestinian generation in the synchronic and diachronic datasets, but it increases significantly in the speech of young (third-generation) Palestinian speakers, who replicate the linguistic conditioning of variant selection in LA. These results bolster the inference of contact-induced change in PA due to the influence of LA. With respect to verbal negation, the findings show that there is convergent change in terms of overall variant rates in this variable system in PA. Evidence suggests that this variable system is undergoing dialect levelling as a result of contact, with socially marked minority variants diminishing over time in the speech of educated Palestinians. The future temporal reference system, however, seems to be less amenable to contact-induced change, despite similarities in surface forms between LA and PA. Results indicate that this variable system is undergoing an internal change in PA independent of contact with LA, which is led by young, educated speakers, in line with what has been observed in PA spoken outside Lebanon (AbuAmsha 2016). Viewed in the aggregate, the results show that even though it is claimed that (morpho-) syntactic variables may be less susceptible to convergent change than phonological variables (Cheshire et al. 2005; Hinskens et al. 2005), we do not find a neat division between phonology and morpho-syntax. Word-medial imala is overtly commented on and explicitly identified by the targeted Palestinian speech community as a marker of Lebanese speech. Its iconic association with Lebanese speech patterns renders it particularly susceptible to long-term dialect accommodation for some Palestinians. Verbal negation is also subject to social evaluation, as gauged from explicit speaker meta-commentary, and socially marked exponents appear vulnerable to attrition over time. By contrast, the expression of the future temporal reference appears less socially indexical than the other variables and is not subject to normative commentary or overt correction. These differences implicate the social salience of the targeted variables as a key factor influencing their susceptibility to convergence. Situating the results in a wider perspective, the findings highlight the utility of the comparative variationist framework in elucidating the process of language change in spoken Arabic, especially in PA as spoken in Beirut, as well as in distinguishing contact-induced change from internally-motivated change. The results of this study indicate that the effects of dialect contact, and critically, the existence of contact-induced change cannot be fully understood without using a multi-faceted comparative approach incorporating horizontal and vertical comparisons. The results converge in demonstrating that an empirically accountable quantitative approach based on actual speech data is capable of transcending the limitations of alternative frameworks of analysis that have been used to investigate change in dialect contact scenarios in the Arabic-speaking world.